Neutrophil‐to‐lymphocyte ratio and D‐dimer are biomarkers of death risk in severe COVID‐19: A retrospective observational study

Abstract Background and Aims Over 4 million deaths from coronavirus disease (COVID)‐19 have been reported in the world. Several biomarkers have been identified that predict disease severity, but there is still a need to identify biomarkers for death risk in severe COVID‐19. We aim to define amongst the biomarkers already identified those which are mostly associated with increased death rate in patients with severe COVID‐19. Methods In this retrospective study conducted in three public hospitals linked to the Medical School of Ribeirão Preto, Brazil, patients with severe COVID‐19 were evaluated regarding biomarkers (neutrophil‐to‐lymphocyte ratio‐NLR, D‐dimer, fibrinogen) of death risk, obtained before administration of corticosteroids. Results Thirty‐nine (32.8%) of the 119 patients included (104 [87.4%] on mechanical ventilation) died during hospitalization. Non‐survivor group had higher median (range) NLR (12.63 [2.6‐115] vs 7.43 [0.43‐31.8]; P = .001), D‐dimer (2.17 [0.27‐20.00] vs 1.57 [0.28‐20.00]; P = .03), but lower fibrinogen (631 [353‐1078] vs 705 [407‐1200]; P = .02). The group with NLR ≥ 10 and D‐dimer ≥ 2 μg/mL had a higher death risk than the group with NLR < 10 and D‐dimer < 2 μg/mL (OR: 5.39; CI 95%: 1.5‐19.42; P = .01). Conclusion High NLR and D‐dimer, especially when combined, are predictors of death risk for patients with severe COVID‐19 and should be incorporated into their evaluation.


| INTRODUCTION
At the end of 2019 a novel coronavirus, known as SARS-CoV-2, was identified in Wuhan, China. 1 Since then, globally, over 200 million confirmed cases of coronavirus disease 2019 (COVID -19) have been reported. In Brazil, since February 2020, more than 20 million people have been diagnosed with COVID-19, of which more than 570 000 died of this disease.
The clinical picture of COVID-19 is characterized by fever, dry cough, dyspnea, fatigue, myalgia, anosmia, and ageusia. 2 Findings in chest radiography or computerized tomography imaging usually show bilateral pulmonary ground-glass opacifications, mainly in posterior and peripheral areas of the lungs. 3 Mortality rate in COVID-19 has been reported to be somewhat less than 1%. 4 The mortality predictors initially identified were older age, elevation of the D-dimer above 1 μg/L and higher SOFA score. 5 Other factors were associated with increased risk of death, for instance, male gender, obesity, cardiovascular disease, diabetes respiratory disease, and ABO blood type. [6][7][8] Several biomarkers have been identified in order to predict the severity of COVID-19, such as lymphocyte count, neutrophil-tolymphocyte ratio (NLR), C-reactive protein (CRP), interleukin-6 (IL-6), and D-dimer. NLR, CRP, and IL-6 reflect inflammation status, whereas D-dimer the coagulation, the two main psychopathological aspects of COVID-19. [9][10][11][12] In this study, we aim to define amongst the biomarkers already identified those which are mostly associated with increased death rate in patients with severe COVID-19 admitted to intensive care unit (ICU).

| Data collection
Epidemiological, demographic, clinical, laboratory, treatment, and outcome data were extracted from electronic medical records using a standardized data collection form. All data were checked by two physicians.
F I G U R E 1 Flow diagram of participants into the study. DD, D-dimer; ICU, intensive care unit; NLR, neutrophil-to-lymphocyte ratio; PCR, real-time polymerase chain reaction

| Laboratory procedures
The method for laboratory confirmation of COVID-19 was the SARS-CoV-2 detection in respiratory specimens by real-time polymerase chain reaction. Routine blood tests were: complete blood count, coagulation profile (including D-dimer and fibrinogen), serum biochemical tests, serum ferritin, lactate dehydrogenase (LDH), homocysteine and C-reactive protein (CRP). The results analyzed were obtained between the date of admission to ICU and before administration of corticosteroids.

| Statistical analysis
Continuous variables were expressed as means and SD or medians and range, according to distribution characteristics (the Kolmogorov-Smirnov test was employed to define distribution characteristics).  Patients' characteristics at ICU admission are detailed in Table 1.
Non-survivor group was older and had a higher rate of blood hypertension and a higher value of SAPS-3 score. The non-survivor group also presented a higher neutrophil-to-lymphocyte ratio (NLR) and Ddimer, but a lower fibrinogen concentration ( Table 2).
In another type of analysis, we divided the participants into two groups according to the NLR cutoff of 10 (median of all patients was 9.29). Similarly, we divided the participants into two groups according to the D-dimer cutoff of 2 (median of all patients was 1.89 μg/mL).
When we combined the two biomarkers and divided the participants into two groups (NLR ≥ 10 and D-dimer ≥ 2 μg/mL vs NLR < 10 and D-dimer < 2 μg/mL) it was observed that the first group had a higher death risk than the second (relative risk: 2.714; CI 95%: 1.516-4.998) ( Table 3).
In the multivariate analysis, it was observed that NLR ≥ 10 and Ddimer ≥ 2 μg/mL were associated with a higher death rate (Table 4).

| DISCUSSION
This retrospective observational study identified perhaps the two most important biomarkers for death risk in patients with severe COVID-19, which are neutrophil-to-lymphocyte ratio and D-dimer. It is important to emphasize that the patients had an extremely severe clinical picture, as approximately 90% of them were on mechanical ventilation and one third died in hospital.
The differences in epidemiological and clinical characteristics between the groups of survivors and non-survivors have already been described in other studies, which identified factors and biomarkers associated with increased death risk. 5,8 In our study, the clinical factors associated with higher mortality, as expected, were older age, blood hypertension and higher value of SAPS-3. As for laboratory evaluation, higher NLR and D-dimer and lower fibrinogen levels on admission were associated with an increased risk of death. Note: SAPS-3 scores were assessed on admission to the study. Abbreviations: BMI, body mass index; SAPS-3 score, simplified acute physiology score 3.
NLR is a marker previously used as a prognostic factor for patients with sepsis, as it is an early indicator of increased systemic inflammation. 10,15 Previous studies reported that increased NLR is found in patients with severe COVID-19 compared to those with non-severe disease, and possibly reflects an immune system dysregulation. [15][16][17][18][19] Jimeno et al observed that patients with severe clinical courses presented higher NLR at admission and higher NLR peak than that observed in non-severe group, and hypothesized that NLR is a marker of endothelial dysfunction. 20 In our study, we identified that NLR is not only important to stratify the severity of the disease, but also to predict mortality in severe cases. 14 days the non-survivor group had lower fibrinogen levels. 25 We found in our study increased fibrinogen levels compatible with severe COVID-19, however, in the non-survivor group fibrinogen values were lower than those found in the survivor group. We hypothesized that in the non-survivor group a progression to a later stage of disseminated intravascular coagulation may be already occurring, with increased consumption of fibrinogen, however, more studies are necessary to investigate this issue.
Hemoglobin concentration was similar in both groups and was within the reference values on admission to ICU. We believe that this finding may be explained by the fact that the decrease of hemoglobin concentration is a later event in an inflammatory picture and our patients here were evaluated relatively early, before a clear impact of inflammation on erythropoiesis. 26 It is important to highlight that patients with NLR ≥ 10 and Ddimer ≥ 2 μg/mL had a significantly higher death risk than the group with values below those defined above. Both parameters combined represent a strong prognostic factor to be considered on admission of patients with severe COVID-19. What is more, beyond their statistical significance, they portray the two most important aspects of COVID-19 pathophysiology: inflammation and coagulation abnormalities.
Fu et al reported that NLR and D-dimer levels were higher in patients with severe COVID-19 compared with the mild/moderate group on days 1, 7, and 14 of the disease. 27 Ye et al found that in hospitalized patients initial and peak D-dimer and NLR were higher in intubated and in deceased patients. 28 Our study adds to the data that combined NLR and D-dimer may not only aid in the differentiation between severe and non-severe COVID-19 as described in the literature, but also predict the death risk in those patients with severe COVID-19.
The main limitation of our study is its retrospective design, carried out in a single region of Brazil. The sample size could be considered relatively small, however, it had a sufficient size to enable us to obtain significant results. We believe that a strong point of our study is that the biomarkers of death risk were evaluated in patients with severe COVID-19, all of them admitted to ICU. D-dimer and, especially, NLR are universally available exams, including in small units of care in lowincome countries. Another strong point of this study is that the participants were recruited from a miscegenated population, which perhaps confers to our findings a universal validity. It is important to emphasize that the NLR should be obtained before introduction of corticosteroids, which knowingly increases neutrophil and decreases lymphocyte counts.
These tests should be incorporated into the general evaluation of death risk for patients with severe COVID-19 because of their clinical significance.

ACKNOWLEDGMENT
The authors thank Nathália Cristine André for her assistance.

CONFLICT OF INTEREST
The authors have no conflict of interest to disclose. Patrícia Oliveira da Cunha Terra had full access to all of the data in this study and takes complete responsibility for the integrity of the data and the accuracy of the data analysis.

DATA AVAILABILITY STATEMENT
The authors confirm that the data supporting the findings of this study are available within the article.

ETHICS STATEMENT
This study was approved by the institutional review board (Comitê de Etica em Pesquisa do Hospital das Clínicas de Ribeirão Preto; CAAE 50193221.7.0000.5440), and was performed in accordance to the Helsinki Declaration of 1964 and its later amendments. As this is a retrospective study, the ethical review board agreed to waiver the signed consent by the patients.