Baseline characteristics of the study population
The study population included 57 hospitalized patients with COVID-19 and 46 hospitalized patients with community acquired pneumonia (CAP). For COVID-19 patients, the median age was 65 years (IQR, 54-72), and 44% were men. For CAP patients, the median age was 64 years (IQR, 60-70), and 59% were men (Table 1). Both of the COVID-19 patients and CAP patients had 1 or more coexisting medical conditions, and compared with COVID-19 patients, CAP patients were more likely to have coexisting medical conditions, including cardiovascular disease (COVID-19 patients vs CAP patients: 4[7%] vs 10[22%]), pulmonary disease (1[2%] vs 16[35%]), and smoking (1[2%] vs 22[48%]) (Table 1).
Table 1 Baseline characteristics of patients on admission
Variable
|
COVID-19 patients(n=57)
|
CAP patients
(n=46)
|
P value
|
Age(years)
|
65(54, 72)
|
64(60, 70)
|
0.471
|
Sex
|
|
|
0.134
|
Men
|
25(44%)
|
27(59%)
|
|
Women
|
32(56%)
|
19(41%)
|
|
Cardiovascular diseases
|
4(7%)
|
10(22%)
|
0.030
|
Pulmonary disease
|
1(2%)
|
16(35%)
|
<0.001
|
Hypertension
|
20(35%)
|
18(39%)
|
0.672
|
Diabetes
|
9(16%)
|
11(24%)
|
0.300
|
Kidney diseases
|
1(2%)
|
3(7%)
|
0.464
|
Smoking
|
1(2%)
|
22(48%)
|
<0.001
|
Fever
|
45(79%)
|
30(65%)
|
0.119
|
Cough
|
24(42%)
|
40(87%)
|
<0.001
|
Shortness of breath
|
26(46%)
|
28(61%)
|
0.123
|
Myalgia
|
3(5%)
|
3(7%)
|
0.786
|
Chest distress
|
21(37%)
|
15(33%)
|
0.654
|
Diarrhoea
|
4(7%)
|
2(4%)
|
0.879
|
Inapptence
|
3(5%)
|
8(17%)
|
0.097
|
Fatigue
|
4(7%)
|
16(35%)
|
<0.001
|
COVID-19 patients: Novel coronavirus pneumonia patients; CAP patients: Community acquired pneumonia; Data are median ± SD, and n (%).
On admission, no matter in COVID-19 patients or CAP patients, most patients had fever, cough, shortness of breath, myalgia, chest distress, diarrhea, inapptence and fatigue. Besides, there were numerous differences in laboratory findings (Table 2). Compared with COVID-19 patients, CAP patients were more likely to have higher white blood cell (WBC) and neutrophil counts (N), as well as higher procalcitonin (PCT), erythrocyte sedimentation rate (ESR) and fibrinogen (FIB), conversely, lower activated partial thromboplastin time (APTT), and there were no significant differences in other biomarkers levels between two groups.
Table 2 Laboratory results of patients with different pneumonia
Variable
|
COVID-19 patients (n=57)
|
CAP patients
(n=46)
|
P value
|
Blood routine
|
|
|
|
WBC (× 10⁹ cells per L)
|
5.5(3.8, 6.9)
|
6.9(5.2, 11.2)
|
0.001
|
N (× 10⁹ cells per L)
|
3.4(2.1, 4.6)
|
5.3(3.7, 9.4)
|
<0.001
|
L (× 10⁹ cells per L)
|
1.3(0.8, 1.6)
|
1.0(0.7, 1.7)
|
0.318
|
PLT (× 10⁹ cells per L)
|
241.0(182.8, 324.0)
|
253.0(174.8, 395.8)
|
0.532
|
HGB (g/L)
|
123.4±16.3
|
112.4±19.0
|
0.001
|
Coagulation function
|
|
|
|
APTT (sec)
|
39.5±6.2
|
36.0±9.1
|
0.033
|
PT (sec)
|
13.8±1.1
|
13.5±2.1
|
0.123
|
TT (sec)
|
17.0(15.9, 18.0)
|
16.7(16.0, 17.5)
|
0.378
|
INR
|
1.1±1.0
|
1.1±0.2
|
0.108
|
FIB (g/L)
|
4.8±1.5
|
5.6±2.0
|
0.014
|
Blood biochemistry
|
|
|
|
Albumin (g/L)
|
34.0±3.9
|
32.5±5.4
|
0.118
|
Globulin (g/L)
|
34.5±5.4
|
31.8±5.9
|
0.011
|
TP (g/L)
|
68.5±4.6
|
64.1±7.5
|
0.001
|
ALT (U/L)
|
21.5(13.3, 37.7)
|
24.3(17.3, 48.7)
|
0.207
|
AST (U/L)
|
26.5(18.8, 34.5)
|
29.1(21.7, 40.5)
|
0.135
|
BUN (mmol/L)
|
4.3(3.3, 5.9)
|
5.1(3.7, 7.1)
|
0.174
|
Cr (μmol/L)
|
67.0(57.0, 82.0)
|
67.2(52.3, 80.8)
|
0.840
|
CK (U/L)
|
67.0(29.0, 178.0)
|
78.1(37.9, 106.9)
|
0.921
|
LDH (U/L)
|
288.8±104.7
|
241.1±67.4
|
0.048
|
Myoglobin (ng/mL)
|
31.4(22.4, 53.5)
|
64.9(42.3, 92.8)
|
0.004
|
Infection-related biomarkers
|
PCT (ng/mL)
|
0.04(0.02, 0.09)
|
0.1(0.05, 0.42)
|
<0.001
|
ESR (mm/h)
|
32.0(18.5, 63.8)
|
72.0(33.0, 98.5)
|
0.001
|
Data are median ± SD, or medians (25th–75th percentile). COVID-19 patients: Novel coronavirus pneumonia patients; CAP patients: Community acquired pneumonia. WBC: White blood cell count; N: Neutrophil count; L: Lymphocyte count; PLT: Platelet count; HGB: Haemoglobin; APTT: Activated partial thromboplastin time; PT: Prothrombin time; TT: Thrombin time; INR: International normalized ratio; FIB: Fibrinogen; TP: Total protein; ALT: Alanine aminotransferase; AST: Aspartate aminotransferase; BUN: Blood urea nitrogen; Cr: Serum creatinine; Creatine kinase; LDH: Lactate dehydrogenase; PCT: Procalcitonin; ESR: Erythrocyte sedimentation rate.
D-dimer levels were related with markers of inflammation
To investigate whether D-dimer levels is associated with levels of inflammatory factors, we performed Spearman’s correlations analysis between D-dimer levels and infection-related biomarkers levels in COVID-19 patients and CAP patients. As shown in Table 3, for COVID-19 patients, D-dimer levels were positively correlated with infection-related biomarkers levels including hsCRP, PCT and ESR before treatments (R=0.426, 0.349, 0.345 respectively, P<0.05). And D-dimer levels also had great correlations with inflammatory cells levels before treatments such as WBC, N, L(R=0.402, 0.464, -0.426, respectively, P<0.01) and coagulation function-related factors levels such as PT, INR(R=0.368, 0.386, respectively, P<0.01). In addition, for CAP patients, there were also positive correlations between D-dimer levels and infection-related biomarkers levels before treatments including hsCRP, PCT (R=0.300, 0.391, respectively, P<0.05), and D-dimer levels were also related with other biomarkers levels before treatments like PT, APTT, INR and CK(R=0.374, 0.383, 0.298, -0.464, respectively, P<0.05). At the same time, we also analyzed the correlations between these indicators after treatments in COVID-19 patients, and found that there were still great correlations between D-dimer and the same biomarkers as above, their correlation coefficients R > 0.3(P<0.05). However, due to the absence of following-up data, we couldn’t analyze these relationships between treated biomarkers levels in CAP patients.
Table 3 Spearman’s correlation coefficients between D-dimer and other biomarkers in COVID-19 patients
D-dimer and
|
Untreated
|
Treated
|
hsCRP
|
0.426**
|
0.495**
|
PCT
|
0.349*
|
0.659**
|
ESR
|
0.345*
|
0.511*
|
WBC
|
0.402**
|
0.325
|
N
|
0.464***
|
0.462**
|
L
|
-0.426**
|
-0.400*
|
PT
|
0.368**
|
0.234
|
APTT
|
-0.056
|
0.123
|
TT
|
-0.016
|
0.103
|
INR
|
0.386**
|
0.194
|
FIB
|
0.282*
|
0.369*
|
CK
|
0.151
|
-0.287
|
hsCRP: Hypersensitive C-reactive protein. The correlations between D-dimer and other biomarkers before(untreated) and after(treated) treatments in COVID-19 patients. *** means P < 0.001 , ** means P < 0.01 , * means P < 0.05 .
Table 4 Spearman’s correlation coefficients between D-dimer and other biomarkers in CAP
patients
D-dimer and
|
Untreated
|
hsCRP
|
0.300*
|
PCT
|
0.391**
|
ESR
|
0.273
|
WBC
|
0.038
|
N
|
0.046
|
L
|
0.006
|
PT
|
0.374*
|
APTT
|
0.383*
|
TT
|
-0.083
|
INR
|
0.398**
|
FIB
|
0.219
|
CK
|
-0.464**
|
The correlations between D-dimer and other biomarkers before(untreated) and after(treated) treatments in CAP patients. *** means P < 0.001 , ** means P < 0.01 , * means P < 0.05 .
More importantly, we found that in COVID-19 patients the correlation between D-dimer levels and hsCRP levels before treatments was related to the levels of hsCRP, while the levels of hsCRP exceed 10mg/l, the correlation between D-dimer and hsCRP was stronger (hsCRP<10mg/l vs hsCRP≥10mg/l, R=-0.212 vs 0.448, Table 5).
Table 5 Spearman’s correlation coefficients between D-dimer and related biomarkers according to untreated hsCRP levels in COVID-19 patients
D-dimer and
|
hsCRP < 10
|
hsCRP ≥ 10
|
hsCRP
|
|
|
Untreated
|
-0.212
|
0.448**
|
Treated
|
-0.268
|
0.348
|
PCT
|
|
|
Untreated
|
-0.178
|
0.320
|
Treated
|
<0.00
|
0.449
|
CK
|
|
|
Untreated
|
0.371
|
-0.067
|
Treated
|
-
|
-0.304
|
*** means P < 0.001 , ** means P < 0.01 , * means P < 0.05 .
D-dimer levels were higher in COVID-19 patients compared with CAP patients on admission
To explore the difference of D-dimer levels between COVID-19 patients and CAP patients, we divided the levels of untreated hsCRP into two groups both in COVID-19 patients and CAP patients, one group for hsCRP levels < 30mg/l, and another group for hsCRP ≥ 30 mg/l. And we found that no matter in COVID-19 patients or CAP patients, the higher hsCRP levels, the higher D-dimer levels (Figure 1A and Figure 2A). Besides, this trend also existed in other related-biomarkers levels, including PCT, FIB and INR (Figure 1B-F and Figure 2B-F). Interestingly, it was worth mentioning that compared with COVID-19 patients, the levels of hsCRP were higher in CAP patients, whereas the levels of D-Dimer were lower in CAP patients (Figure 3A-B).
In COVID-19 patients with good clinical prognosis, hsCRP levels decreased after treatment, while D-dimer levels decreased synchronously
As previous described, D-dimer levels were truly related with biomarkers of inflammation, especially with hsCRP. We then analyzed the specific relationship between D-dimer levels and hsCRP levels in COVID-19 patients, and found that both hsCRP levels and D-dimer levels decreased after treatments (Figure 4A-B). Moreover, we analyzed their relationship before and after treatments stratified by untreated hsCRP quartiles, as expected, after therapy, hsCRP levels were significantly decreased in the 2nd, 3rd and 4th quartiles of untreated hsCRP (Figure S1 A-D), and there were also a downward trend in D-dimer levels at the different quartiles (Figure S1 E-H).
However, considering that the values stratified by untreated hsCRP quartiles might be higher or lower cutoff values which could bias the results, and as previously described, hsCRP levels were significantly decreased in the 2nd quartile, we then divided all patients into two groups based on the cutoff value 10mg/l of untreated hsCRP levels: hsCRP < 10 mg/dl, hsCRP ≥ 10 mg/dl. Obviously, after treatments, the decrease of D-dimer levels was synchronous with the decrease of hsCRP levels (Figure 4C-F).
In addition, it’s worth mentioning that there were 53 patients were cured or turned into mild cases, whereas 4 patients were died in our study. More important, we found that in deceased patients, both the untreated hsCRP or D-dimer levels and treated hsCRP or D-dimer levels were still abnormally high (Table S1), conversely, both hsCRP and D-dimer levels significantly decreased in patients with a good clinical prognosis after therapy.
In COVID-19 patients, some patients had a significant decrease in hsCRP levels after therapy, whereas D-dimer levels were increased
In our study, both D-dimer levels and hsCRP levels decreased after treatments in COVID-19 patients. In order to investigate whether the levels of D-dimer also decreased significantly in those patients with a significant decrease in hsCRP levels, we then analyzed the relationship between the extent of decline in hsCRP and D-dimer levels after treatments. Interestingly, it was worth mentioning that some patients had a significant decrease in hsCRP levels, whereas their D-dimer levels were increased (Figure 5), highlighting the possibility for aggressive coagulation therapy. Therefore, for these patients, the anticoagulant therapy was strengthened, and the low molecular weight heparin was changed from the preventive dose to the therapeutic dose.