DOI: https://doi.org/10.21203/rs.3.rs-2196905/v1
As a trigger of acute myocardial infarction (AMI), bacterial infection usually accompanies with AMI, and will lead to worse outcomes of AMI patients. We aimed to assess the prognosis of AMI with concomitant bacterial infection and find out the best laboratory examinations to recognize it.
All patients hospitalized for an AMI in cardiology department were prospectively included. Patients were stratified into those with or without concomitant bacterial infection. Outcomes and laboratory examinations were compared between groups in unadjusted and adjusted analyses.
Among the 456 patients hospitalized for AMI, 120 (26%) had a concomitant diagnosis of bacterial infection. Out-comes in hospital were worse in patients with bacterial infection (more acute heart failure: 61.7% vs. 22.6%, p < 0.001, and higher all-cause mortality in hospital: 15% vs. 3.9%, p < 0.001). In the Receiver Operating Curves (ROC) of biomarkers of AMI and concomitant bacterial infection, Areas under the Receiver Operating Curves (AUC) for c-reaction protein (CRP) and CRP to Platelet Ratio (CRP/PLA) were higher than Neutrophil to Lymphocyte Ratio (NLR) and leucocyte count. (0.852(0.81–0.89) mg/L, 0.848(0.81–0.89) *10− 9mg, p < 0.001). The sensitivity of CRP and the Specificity of CRP/PLA were the highest (80% and 88%). After adjusting for confounders, CRP/PLA (> 0.08) was associated with a fivefold increased risk of bacterial infection when compared with other biomarkers (OR (95%CI) = 5.62 (2.64–11.96), p < 0.001). CRP (> 8.05) was also associated with a higher risk of bacterial infection (OR (95%CI) = 4.02 (1.81–5.85), p = 0.001).
Bacterial infection will lead to worse outcomes of AMI patients, including in-hospital mortality and heart failure. It is the first time to use the CRP/PLA to distinguish AMI with concomitant bacterial infection from other AMI patients. The combination of CRP > 8.05 mg/L and CRP/PLA > 0.08*10− 9mg was the best hallmark of AMI with concomitant bacterial infection.
Globally, acute myocardial infarction (AMI), which is caused most often by atherosclerosis, is the leading cause of death. Atherosclerosis is a chronic disease characterized by systemic lipid accumulation and immune inflammation[1]. AMI is associated with leukocytosis and monocities throughout the acute inflammation and regenerative process following cardiac or vascular injury. The inflammatory response is also reflected in C-reactive protein (CRP) and Neutrophil to Lymphocyte Ratio (NLR), which are positive in most AMI patients.
Recent studies have shown that influenza, pneumonia, acute bronchitis, and other chest bacterial infections increase the short-term risk of myocardial infarction[2–5]. AMI risk increases six times, four times, and three times after bacterial infections with influenza virus, respiratory syncytial virus, or other respiratory viruses[2]. The bacterial infection that related to AMI also include urinary tract bacterial infection[3] and bacteremia[6]. The potential mechanisms include the increasing of inflammatory activity in atheromatous plaques after an infectious stimulus[7], the augment of the risk of coronary thrombosis at sites of plaque in the prothrombotic, procoagulant state that is associated with acute bacterial infection[8, 9], the up regulation of the metabolic needs of peripheral tissues and organs[10], and the cardiac lesions that are characterized by vacuolization and loss of myocytes without accumulation of inflammatory cells caused by pneumococcal bacteremia[11]. Patients with AMI usually manifest fever at admission, as well as elevation of CRP and leucocytes counts, but it does not necessarily mean that an bacterial infection is present[12, 13]. Fever and bacterial infection biomarkers are difficult to distinguish between the AMI, a complicating bacterial infection, and other causes of symptoms
In this study of acute myocardial infarction and concomitant bacterial infection patients from all AMI in cardiology department, we aimed to assess the prognosis of AMI with concomitant bacterial infection and find out the best laboratory examinations to recognize it.
This study was approved by the Ethics Committee of Wenzhou Medical University and was in compliance with the Declaration of Helsinki. All patients provided written consent prior to participation. We have consecutive studied 120 acute myocardial infarction and concomitant bacterial infection patients admitted to the cardiology department of The Second Affiliated Hospital of Wenzhou Medical University between January 1, 2019 and November 30, 2019 and 336 sex matched AMI subjects selected from the same department (non-bacterial infection cases).
There was a rise in Troponin-I values and at least one of the following features were present in the diagnosis of AMI: A new ischemic ECG change; pathological Q waves; symptoms of myocardial ischemia; Imaging evidence of new loss of viable myocardium or new regional wall motion abnormalities consistent with ischemic etiology; coronary thrombus detected by angiography or CT [14]. AMI with concomitant bacterial infection was defined that the diagnosis of acute bacterial infection had been made when the onset of AMI symptoms. Acute bacterial infection was defined by a physician's diagnosis of signs and symptoms of infection
At admission, we collected demographic, clinical, and biological information as well as ECGs and other tests. Within 24 hours of admission, blood samples were collected to measure CRP, leukocyte count, neutrophil count, lymphocyte count, and platelet count. Troponin-I was measured within 20 minutes after admission.
We collected data on all-cause hospital deaths (e.g., cardiogenic shock, ventricular rhythm disorders, sepsis shock, sudden unexpected deaths). The concept of heart failure was defined by the maximal Killip class 2–4 (i.e., acute pulmonary edema and cardiogenic shock).
Data were presented as means*standard deviations or medians and interquartile ranges for continuous variables. In this study, continuous variables were compared using Student's t-tests or Mann-Whitney tests, while dichotomous variables were compared using Chi2 or Fisher's tests. Logistic regression model was used to compare concomitant bacterial infection patients with other AMI patients, including the relevant variables associated with bacterial infection and mortality. All analyses were conducted using SPSS version 17.0 (IBM Inc., Armonk, NY, USA). The significance level was set at 0.05.
Among the 456 patients hospitalized for AMI, 120 (26%) had a concomitant diagnosis of bacterial infection. The baseline characteristics are presented in Table 1. Patients with concomitant bacterial infection were 10 years older than other AMI patients (median age 75 (64–83) vs. 65 (55–75) y, p < 0.001), and had more chronic renal failure history (37% vs. 15%, p < 0.001), more cardiogenic shock at admission (16% vs. 7%, p < 0.001), but less hypercholesterolemia (34% vs. 47%, p = 0.016). Out-comes in hospital were worse in patients with bacterial infection (i.e., more acute heart failure: 61.7% vs. 22.6%, p < 0.001, and higher all-cause mortality in hospital: 15% vs. 3.9%, p < 0.001). In addition, troponin I at admission was much higher in bacterial infection group (10.7 (1.6–34) vs. 2.5 (0.2–18.7) ng/mL). As expected, C-reactive protein level was much higher in the group with AMI and concomitant bacterial infection (36.3 (9.2–102.8) vs. 3.16 (1.6–8.4) mg/L). So as to the leucocyte count (10.55 (7.6–13.4) vs. 8.83 (6.9–11.2) *109/L) and the neutrophil count (8.46 (6.0–11.3) vs. 6.5 (4.6–9.3) *109/L). Whereas the lymphocyte count was lower in the bacterial infection group (1.1(0.8–1.61 vs. 1.5(1.0–2.0) *109/L).
|
Infection |
Noninfected |
P |
|
|---|---|---|---|
|
n = 120 |
n = 336 |
||
|
Risk factors and comorbidities |
|||
|
Age, years |
75(64–83) |
65(55–75) |
< 0.001 |
|
Female |
39(33%) |
84(25%) |
0.112 |
|
Hypertension |
82(68%) |
212(63%) |
0.303 |
|
Hypercholesterolemia |
40(34%) |
150(47%) |
0.016 |
|
Diabetes |
47(39%) |
107(32%) |
0.130 |
|
Chronic renal failure |
44(37%) |
54(15%) |
< 0.001 |
|
Anemia |
7(6%) |
7(2%) |
0.086 |
|
Clinical data at admission |
|||
|
HR, beats/min |
88(76–105) |
78(68–90) |
0.001 |
|
SBP, mmHg |
123(110–145) |
133(119–150) |
< 0.001 |
|
DBP, mmHg |
73(64–84) |
80(73–90) |
< 0.001 |
|
Cardiogenic shock |
19(16%) |
23(7%) |
0.004 |
|
Temperature, ℃ |
37.7(37.2–38.2) |
37.3(37-37.6) |
< 0.001 |
|
ECG at admission |
|||
|
STEMI |
73(61%) |
179(53%) |
0.153 |
|
Biological data |
|||
|
CRP, mg/L |
36.3(9.2-102.8) |
3.2(1.6–8.6) |
< 0.001 |
|
Leucocyte, *109/L |
10.6(7.6–13.4) |
8.8(6.9–11.2) |
0.001 |
|
Neutrophil, *109/L |
8.5(6.0-11.3) |
6.5(4.6–9.3) |
< 0.001 |
|
Lymphocyte, *109/L |
1.1(0.8–1.6) |
1.47(1.0–2.0) |
< 0.001 |
|
Platelet, *109/L |
206.0(157.0-255.0) |
204.5(173.4-251.7) |
0.147 |
|
Troponin I admission, ng/mL |
10.7(1.6–34.0) |
2.50(0.2–18.7) |
< 0.001 |
|
Acute management |
|||
|
Early vascularization |
91(76%) |
293(87%) |
0.002 |
|
Hospital outcomes |
|||
|
All cause death |
18(15%) |
13(3.9%) |
< 0.001 |
|
Heart Failure |
74(61.7%) |
76(22.6%) |
< 0.001 |
| CRP: C-reactive protein; DBP: diastolic blood pressure; HR: heart rate; IQR: interquartile range; SBP: systolic blood pressure; STEMI: ST segment elevation myocardial infarction. | |||
All-cause in-hospital death was four times as common in bacterial infection group (15% vs. 3.9%, p < 0.001). And heart failure was nearly three times (61.7% vs. 22.6%, p < 0.001) In patients with concomitant bacterial infection, heart failure was nearly twice times more frequent in the age ≥ 75 group (75.4% vs. 45.5%, p = 0.001). But all-cause mortality was no significant difference between two groups (18.5% vs. 10.9%, p = 0.248).
The Receiver Operating Curves (ROC) for biomarkers of AMI and concomitant bacterial infection are presented in Fig. 1, including CRP, CRP/PLA, NLR and leucocyte count. The AUC for CRP and CRP/PLA were higher than two others (0.852(0.81–0.89) mg/L, 0.848(0.81–0.89) *10− 9mg, p < 0.001. The sensitivity of CRP and the Specificity of CRP/PLA were the highest (80% and 88%).
Each biomarker was divided into two groups depended on the cutoff value in the ROC (Fig. 1). After adjustment for sex, age, ECG at admission and some risk factors for AMI such as hypertension, hypercholesterolemia, diabetes, chronic renal failure, CRP/PLA(> 0.08) was associated with a fivefold increased risk of bacterial infection when compared with other biomarkers (OR (95%CI) = 5.62 (2.64–11.96), p < 0.001). CRP(> 8.05)was also associated with a higher risk of bacterial infection(OR (95%CI) = 4.02 (1.81–5.85), p = 0.001) (Table 2). In the subgroup of patients younger than 75 (n = 306), this association of CRP/PLA(> 0.08) was even greater after adjustment (OR (95%CI) = 8.65 (3.05–24.48), P < 0.001) (Table 3).
|
Univariable |
Mode 1 |
|||
|---|---|---|---|---|
|
OR (95% CI) |
P |
OR (95% CI) |
P |
|
|
Leucocytes |
2.35(1.54–3.60) |
< 0.001 |
2.00(1.06–3.78) |
0.033 |
|
NLR |
3.53(2.25–5.53) |
< 0.001 |
3.09(1.63–5.50) |
0.001 |
|
CRP |
11.63(6.98–19.36) |
< 0.001 |
4.02(1.81–5.85) |
0.001 |
|
CRP/PLA |
16.29(9.78–27.12) |
< 0.001 |
5.62(2.64–11.96) |
< 0.001 |
| Mode 1: adjusted for sex, age, ECG at admission, hypertension, hypercholesterolemia, diabetes and chronic renal failure. | ||||
| CI: confidence interval; OR: odds ratio | ||||
|
Age＜75 |
Age ≥ 75 |
||||||
|---|---|---|---|---|---|---|---|
|
OR (95% CI) |
P |
OR (95% CI) |
P |
||||
|
Leucocyte |
2.57(1.04–6.36) |
0.041 |
1.75(0.67–4.61) |
0.257 |
|||
|
NLR |
2.90(1.16–7.25) |
0.023 |
3.48(1.48–8.17) |
0.004 |
|||
|
CRP |
3.79(1.31–10.95) |
0.014 |
4.18(1.28–13.64) |
0.018 |
|||
|
CRP/PLA |
8.65(3.05–24.48) |
< 0.001 |
3.45(1.05–11.31) |
0.041 |
|||
| CI: confidence interval; OR: odds ratio; NLR: Neutrophil to Lymphocyte Ratio; CRP: C-reaction protein; CRP/PLA: C-reaction protein to Platelet Ratio. | |||||||
In this study of patients with AMI and concomitant bacterial infection, the main findings were as follows: (1) In-hospital deaths and heart failure were significantly more common in AMI patients with concomitant bacterial infection compared with those without bacterial infection. (2) It is the first time to use the CRP to platelet ratio to distinguish AMI with concomitant bacterial infection from other AMI patients. The combination of CRP > 8.05 mg/L and CRP/PLA > 0.08*10− 9mg was the best hallmark of AMI complicating bacterial infections.
The leading cause of death worldwide is AMI, a systemic, lipid-driven immune inflammatory disease. Acute bacterial infections such as influenza, pneumonia, acute bronchitis, urinary tract bacterial infection and bacteremia[2–5] usually induce or accompany AMI and lead to worse outcomes and longer hospital stays. Based on previous studies, Serious bacterial infection of AMI patients was not only resulted in longer hospital stay, but also associated with significantly higher rates of 90-day death[15]. Among the possible mechanisms is an increase in inflammation in atheromatous plaques during infection [7]. When an acute bacterial infection causes a prothrombotic, procoagulant state, the risk of coronary thrombosis increases [8, 9]. The up regulation of the metabolic needs of peripheral tissues and organs[10], and the cardiac lesions that are characterized by vacuolization and loss of myocytes without accumulation of inflammatory cells caused by pneumococcal bacteremia[11].
In our research, a 4 times increased risk of in-hospital mortality of concomitant bacterial infection subjects compared with other AMI patients, and 3 times of heart failure. The percentage of heart failure in AMI with concomitant bacterial infection patients gradually increasing with age, while the mortality was no difference between age༜75 and age ≥ 75 groups maybe because of the small sample size of deaths in these two groups. After adjusting confounders (age, gender, hypertension, diabetes, anemia and some risk factors), AMI with concomitant bacterial infection was no relationship with all-cause mortality and the mortality of shock was 12 times compared to other groups. The effect of bacterial infection on mortality maybe occurred through cardiogenic shock. It is responsible for concomitant bacterial infection to a worse prognosis of AMI patients. Healthcare providers must acknowledge these risks and not dismiss the fever and the elevation of bacterial infection biomarkers as trivial events in patients with AMI. It is an urgent need to determine whether the fever and the elevation of bacterial infection biomarkers are due to the MI, to a concomitant bacterial infection, or to other causes of systemic inflammation when an AMI patient is at admission.
Leukocytes is one of the markers of bacterial infection. Acute bacterial infections are characterized by mature and immature neutrophils. But leukocytosis is only a suggestion but not a diagnosis of bacterial infection. According to the Van den Brule A’s research, the sensitivity and specificity of leukocytosis in diagnosing acute appendicitis are 62% and 75%, respectively[16, 17]. Leukocytes also play an important role in the pathophysiology of AMI, such as permeating endothelial cells, inducing the formation of micro vascularity there, and making plaques more susceptible to rupture[18]. According to the research of Hilde E. et, al, leukocyte count increased across the coronary vascular decease continuum, which mainly depended on the increase in neutrophil count[19].
A hyperactive CRP is detected during inflammatory illnesses like rheumatoid arthritis, cardiovascular disease, and bacterial infection [20, 21]. According to Healy and Freedman’s research, CRP level increased in all bacterial infection cases compared to non-bacterial infection cases. But there is no differences between the bacterial infection types[22]. So, CRP is a marker of bacterial infection and inflammation. Since that the atherosclerosis is a systematic chronic inflammatory process, it may result in detectable levels of acute phase reactants and pro-inflammatory cytokines in the blood. Myocardial infarction and unstable angina are associated with elevated CRP; high levels of these factors predict a poor outcome [23–25].
From what has been discussed above, most of the biomarkers of bacterial infection are also present high levels in AMI. So, one of the main interests of our work was to find out the best biomarker to distinguish AMI patients with concomitant bacterial infection from other AMI patients. It is the first time to use the CRP to platelet ratio (CRP/PLA) to distinguish it. Our results showed that the AUC of CRP and CRP/PLA were over 0.8. The sensitivity of CRP and the specificity of CRP/PLA were highest. The logistic regression also revealed that the two biomarkers associated with a higher risk of bacterial infection. The combination of CRP > 8.05 mg/L and CRP/PLA > 0.08*10− 9mg was the best hallmark of AMI with concomitant bacterial infection. Currently, there is no published data to confirm our results, and further prospective studies are clearly needed to better understand CRP/PLA.
Several limitations have to be acknowledged. Firstly, limited data concerning acute bacterial infection and coronary angiography characteristics were collected, as about 15% patients did not undergo coronary angiography and the symptoms of bacterial infection were not typical. Secondly, further prospective studies are needed to address the relative risk (RR) of CRP/PLA in AMI with concomitant bacterial infection. Lastly, further classification of bacterial infection should be made to detect the value of CRP/PLA in different kinds of bacterial infection.
Bacterial infection will lead to worse outcomes of AMI patients, including in-hospital mortality and heart failure. It is the first time to use the CRP/PLA to distinguish AMI with concomitant bacterial infection from other AMI patients. The combination of CRP > 8.05 mg/L and CRP/PLA > 0.08*10-9mg was the best hallmark of AMI with concomitant bacterial infection.
Data Availability
Upon request, the corresponding author can provide the excel data used to support the findings of this study.
Authors' contributions
Lu Qian designed the overall study with contributions from Zhuoyuan Li. Zhuoyuan Li designed and collected and analyzed data, and cowrote the paper. Suqin Zhang designed, collected and analyzed data with Zhuoyuan Li and Lu Qian. Suqin Zhang, Jiafeng Lin, Zhuoyuan Li and Lu Qian discussed and edited the paper. Jiafeng Lin supervised this study, designed and analyzed data, as well as writing the paper.
Funding
not applicable