Evaluation of Biomarkers of Severity in Patients with COVID-19 Infection

Object: Although many Japanese patients infected with coronavirus disease 2019 (COVID-19) only experience mild symptoms, in some cases a patient’s condition deteriorates, resulting in a poor outcome. This study examines the behavior of biomarkers in patients with mild to severe COVID-19. Methods: The disease severity of 152 COVID-19 patients was classified into mild, moderate I, moderate II, and severe, and the behavior of laboratory biomarkers was examined across these four disease stages. Results: The median age and male/female ratio increased with severity. The mortality rate was 12.5% in both moderate II and severe stages. Underlying diseases, which were not observed in 45% of mild stage patients, increased with severity. An ROC analysis showed that C-reactive protein (CRP), ferritin, procalcitonin (PCT), hemoglobin (Hb) A1c, albumin, and lactate dehydrogenase (LDH) levels were significantly useful for the differential diagnosis of mild/moderate I stage and moderate II/severe stage. In the severe stage, Hb levels, coagulation time, total protein, and albumin were significantly different on the day of worsening from those observed on the day of admission. The frequency of hemostatic biomarker abnormalities was high in the severe disease stage. Conclusion: The evaluation of severity is valuable, as the mortality rate was high in the moderate II and severe stages. The levels of CRP, ferritin, PCT, albumin, and LDH were useful markers of severity, and hemostatic abnormalities were frequently observed in patients in the severe disease stage.


Materials and Methods
One hundred and fifty-two patients with COVID-19 infection (median age, 53.5 years; 25-75th percentile, 38.0-70.0 years; female, n = 78; male, n = 74) were admitted to Mie Prefectural General Medical Center from 1 April 2020 to 21 March 2021. The severity of COVID-19 was evaluated based on the Japanese Medical Care Guidelines [15]. There were 63 patients with mild disease (without pneumonia), 48 with moderate I disease (pneumonia without oxygen therapy), 32 with moderate II disease (pneumonia with oxygen therapy without mechanical ventilation), and 9 with severe disease (pneumonia with mechanical ventilation) ( Table 1).

Statistical Analyses
The data are expressed as the median (25-75th percentiles). The difference in the frequency was analyzed by the chi-squared test. The significance of differences between groups was examined using the Mann-Whitney U test. p-Values < 0.05 were considered to indicate statistical significance. Cut-off values were determined by a receiver operating characteristic (ROC) analysis. A multivariate analysis with a stepwise regression test was performed. All statistical analyses were performed using Stat-Flex software (version 6; Artec Co. Ltd., Osaka, Japan).

Results
The median age and male/female ratio increased with severity ( Table 1). The mortality rate was 12.5% in both the moderate II and severe stages. Comorbidity, which was not observed in 45% of mild stage patients, increased with severity (Table 1).
Hypertension, hyperlipidemia, or diabetes mellitus were observed ≥40% of patients with moderate II or severe disease. Heart failure, other pneumonia complications, and cerebrovascular accident were frequently observed among the patients who died within 30 days.
Regarding the laboratory data (Table 2), the neutrophil, lymphocyte, and platelet counts as well as the PT-INR and D-dimer levels in the moderate II and severe stages were significantly different from those in the mild and moderate I disease stages. The CRP, PCT, and ferritin levels in the moderate II and severe stages were significantly higher than those in the mild and moderate I disease stages. In the moderate II and severe stages the total protein and albumin levels, A/G ratio, and estimated glomerular filtration rate (eGFR) were significantly lower, and the levels of AST, LDH, and hemoglobin A1C levels were significantly higher in comparison with the mild and moderate I disease stages.
The ROC analysis of laboratory biomarkers in the mild and moderate I stages vs. moderate II and severe stages (Table 3, Figure S1), CRP (Figure 1), albumin, and A/G ratio showed ≥80.0% sensitivity; CRP, ferritin, albumin, A/G ratio, and hemoglobin A1c showed an odds ratio of ≥10.00; and CRP, PCT, ferritin, albumin, A/G ratio, and LDH showed an area under the curve (ARC) ≥0.800. In our multivariate analysis, CRP (p = 0.006) and AST (p = 0.033) were identified as significant factors in mild to moderate II stage vs. severe stage; albumin (p = 0.022), CRP (p = 0.001), and LDH (p = 0.011) were identified as significant factors in mild to moderate I stage vs. moderate II to severe stage; and LDH (p = 0.032) and CRP (p = 0.036) were identified as significant factors in mild stage vs. moderate I to severe disease stages.  WBC, white blood cell; PT-INR, prothrombin time-international normalized ratio; APTT, activated partial thromboplastin time; CRP, C reactive protein; A/G ratio, albumin/globulin ratio; T-Bil, total bilirubin; LDH, lactate dehydrogenase; eGFR, estimated glomerular filtration rate; Mild, without pneumonia; Moderate I, pneumonia without oxygen therapy; Moderate II, pneumonia with oxygen therapy; Severe, pneumonia with mechanical ventilation. *, p < 0.05; **, p < 0.01; ***, p < 0.001 in comparison with mild stage. # , p < 0.05; ## , p < 0.01; ### , p < 0.001 in comparison with moderate I. and AST (p = 0.033) were identified as significant factors in mild to moderate II stage vs. severe stage; albumin (p = 0.022), CRP (p = 0.001), and LDH (p = 0.011) were identified as significant factors in mild to moderate I stage vs. moderate II to severe stage; and LDH (p = 0.032) and CRP (p = 0.036) were identified as significant factors in mild stage vs. moderate I to severe disease stages.   In patients with severe disease, the neutrophil, hemoglobin, and platelet counts; PT-INR, APTT, TP, and albumin levels; and the A/G ratio on the day of each instance of worsening were significantly different from those on the day of admission ( Table 4). The frequencies of hemostatic abnormalities were as follows: hemoglobin ≤ 10 g/dL, 44.4%; PT-INR ≥ 1.26, 44.4%; D-dimer ≥ 3.0 µg/mL, 44.4%; platelet count ≤ 10 × 10 10 /L, 55.5%; and APTT ≥ 50 s, 77.8% (Table 5). Table 4. Laboratory data (data on admission vs. data at time of worsening) in patients with severe stage disease.

Admission Data
Worsening

Discussion
In the present study, the mortality rate was 0% in the mild and moderate I disease stages of COVID-19 infection, 12.5% in both moderate II and severe disease stages, and 3.3% overall, suggesting a mortality rate similar to other reports [3,16]. These findings suggest that four patients without mechanical ventilation died, and the direct cause of death was not COVID-19 in these patients. In addition to the management of severe COVID-19, the prevention of progression from mild/moderate I to moderate II and severe disease stages is important. Comorbidity, diabetes mellitus, hypertension, and cerebral vascular accident were related to the disease severity of COVID-19 infection. The elevation of Hb A1c and onset of hypertension may have occurred before the onset of COVID-19 infection, suggesting the importance of antihypertensive therapy and glycemic control before the onset of COVID-19 infection.
An increased WBC, especially increased neutrophil counts and decreased lymphocyte counts, was useful for predicting the disease severity of COVID-19. A decrease in lymphocytes due to apoptosis of CD4+ lymphocytes is correlated with the elevation of inflammatory cytokines [17]. CRP and ferritin were also useful biomarkers for predicting the disease severity of COVID-19. Ferritin elevation is believed to be caused by the cytokine storm and secondary hemophagocytic lympho-histiocytosis [18]. Elevation of CRP and ferritin and a reduction in lymphocytes may occur in association with a cytokine storm [19]. Although there was a large difference in the PCT level, this level was within normal range in the mild, moderate I, and moderate II disease stages. PCT is a specific biomarker for bacterial infection and is not markedly elevated in patients with common viral infections [20]. Therefore, PCT elevation may suggest bacterial co-infection or markedly severe inflammation in the severe disease stage of COVID-19 infection.
The change in AST was small and levels fell to within the normal range in patients with moderate II disease. LDH levels have been reported to be associated with Acute Physiology and Chronic Health Evaluation (APACHE) II, Sequential Organ Failure Assessment (SOFA), and computed tomography (CT) semiquantitative rating scores [21], and LDH is suggested to be a useful marker for multiple organ failure. TP and albumin levels, A/G ratio, and eGFR are also related to multiple organ failure. Thus, these biomarkers are associated with organ failure or renal function, indicating a correlation with the disease severity of COVID-19 infection.
Although D-dimer is a well-known biomarker of thrombosis in COVID-19 [22], the increase of D-dimer was small in the present study. These findings suggest that in the present study there were a few patients with severe disease with no resulting thrombotic complications. In the severe disease stage, Hb and platelet counts were significantly decreased, and the PT-INR and APTT were significantly prolonged, suggesting that the hemostatic abnormalities occurred after admission in patients with severe COVID-19. A high frequency of hemostatic abnormalities, including Hb, platelet count, PT-INR, and Ddimer abnormalities, suggests that a coagulopathy similar to thrombotic microangiopathy (TMA) [23] with disseminated intravascular coagulation (DIC) [24] may occur in patients with severe COVID-19. DIC and TMA produce similar pathological states and are associated with poor outcomes due to marked platelet activation [25]. Accordingly, aspirin treatment was reported to improve the mortality of patients with COVID-19 [26].
There are several limitations associated with the present study. Specifically, this study was observational and thus affected by comorbidities and varying therapies. Several patients died of old age or due to comorbidities rather than due to the severity of their COVID-19 infection. For most patients without severe disease, blood sampling was performed on the day of admission. In addition, the study population was relatively small, and there were few cases in the severe stage.

Conclusions
The mortality rate was high in COVID-19 patients with moderate II/severe disease. CRP, ferritin, PCT, albumin, and LDH levels were useful markers of severity, and hemostatic abnormalities were frequently observed in patients with severe disease.