| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
Ann Thorac Surg 2006;82:2170-2178
© 2006 The Society of Thoracic Surgeons
Division of Cardiac Surgery, Dipartimento d’Emergenza e Trapianti d’Organo, University of Bari, Bari, Italy
Accepted for publication June 15, 2006.
* Address correspondence to Dr Paparella, Division of Cardiac Surgery, Dipartimento d’Emergenza e Trapianti d’Organo, University of Bari, Piazza Giulio Cesare 11, 70100 Bari, Italy (Email: paparella{at}tin.it).
 |
Abstract |
|---|
 Top Abstract IntroductionMaterial and MethodsResultsCommentReferences |
|---|
METHODS: From August 2000 to May 2004, values for preoperative CRP were available for 597 unselected patients undergoing cardiac operations. CRP was used to divide this cohort in two groups: a low inflammatory status (LHS) group of 354 patients with CRP of less than 0.5 mg/dL, and a high inflammatory status (HIS) group of 243 patients with a CRP of 0.5 mg/dL or more. Follow-up lasted a maximum of 3 years (median, 1.8 ± 1.5 years) and was 92.6% complete.
RESULTS: In-hospital mortality was 8.2% in the HIS group and 3.4% in the LIS group (odds ratio [OR], 2.61; p = 0.02). Incidence of postoperative infections was 16.5% in the HIS group and 5.1% in the LIS group (OR, 3.25; p = 0.0001). Sternal wound infections were also more frequent in the HIS group (10.7% versus 2.8%; OR, 3.43; p = 0.002). During follow-up, the HIS group had worse survival (88.5% ± 2.9% versus 91.9% ± 2.5%; OR, 1.93; p = 0.05) and a higher need of hospitalization for cardiac-related causes (73.6% ± 6% versus 86.5% ± 3.2%; OR, 1.82; p = 0.05).
CONCLUSIONS: Patients undergoing cardiac surgery with a CRP level of 0.5 mg/dL or more are exposed to a higher risk of in-hospital mortality and postoperative infections. Despite surgical correction of cardiac disease, a high preoperative CRP value is an independent risk factor for mid-term survival and hospitalization for cardiac causes.
|
Introduction |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
In addition to coronary atherosclerosis, CRP levels can be also high in patients with other heart diseases. CRP is elevated in patients with calcific aortic stenosis [10, 11], which has been recognized as an inflammatory disease with some histologic analogies with coronary atherosclerosis [12]. High plasma concentrations of CRP are also common in patients with chronic rheumatic valve disease [13]. Moreover, CRP can significantly predict the outcome after percutaneous balloon mitral valve commissurotomy [14]. Patients with heart failure, regardless of the cause, have a chronic activation of the inflammatory system that is probably supported by the renin-angiotensin-aldosterone system [15], and relationships between elevated CRP, New York Heart Association class, rehospitalization rates, and mortality were recently observed in these patients [16].
Few studies have analyzed the effects of preoperative inflammatory status on the operative outcome of patients that have undergone cardiac surgery [17–20]. Most of these studies showed that patients with a preoperative inflammatory status had higher incidence of postoperative infections. Yet, the role of the preoperative inflammatory status on mid-term outcome of patients who have undergone cardiac surgery remains unknown.
|
Material and Methods |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
CRP values were used to divide this cohort in two groups: a low inflammatory status (LIS) group of 354 patients with CRP levels of less than 0.5 mg/dL, and a high inflammatory status (HIS) group of 243 patients with CRP levels of 0.5 mg/dL or more. This cut-off has been previously used to assess the risk of cardiovascular events in different populations [3, 7, 18, 20].
The operations performed are listed in Table 1. In 122 patients (47 HIS, 75 LIS), CRP values were also available on the day of operation and postoperative days 1, 2, and 4. The perioperative characteristics of this subgroup of patients were compared with those of the entire population, and no differences were found.
|
The primary end point of this study was to evaluate the effect of preoperative inflammatory status on mid-term survival and hospitalization for cardiac causes after cardiac surgery. The secondary end point was to verify the effect of preoperative inflammatory status on in-hospital mortality and morbidity.
Operative Technique
After premedication with lorazepam, anesthesia was induced with a combination of fentanyl, midazolam, and sodium thiopental and was maintained with propofol. The surgical approach was always a median sternotomy. Heparin was given (300 U/Kg) and CPB was performed at moderate hypothermia (34°C). Cardiac arrest was induced and maintained with cold blood cardioplegia.
Intraoperative heparin monitoring was performed by standard activated clotting time. An additional heparin bolus (5000 U) was given if the activated clotting time was less than 400 seconds. To reverse heparin, protamine was administered at the rate of 1 mg for each milligram of total heparin given before and during CPB. Tranexamic acid was administered in all patients undergoing CPB.
Off-pump coronary revascularizations were done with the use of Lima stitch and intracoronary shunts in all patients. Anesthesia and anticoagulation management was the same as that used for patients who underwent CPB.
In all patients undergoing valve replacement, mechanical bileaflet prostheses were implanted.
Definitions
Systemic hypertension was defined as blood pressure exceeding 140/90 mm Hg, a history of high blood pressure, or the need for antihypertensive medications. Patients were considered to be diabetic if they had history of diabetes mellitus, regardless of the duration of disease, or needed antidiabetic agents. Hypercholesterolemia was defined as a fasting cholesterol level exceeding 200 mg/dL. Smoking history was defined as any current or past form of tobacco use. Chronic obstructive pulmonary disease (COPD) was defined as forced expiratory volume in 1 second of less than 75% or the need of pharmacologic therapy for the treatment of chronic pulmonary compromise. On-going refractory angina that required the use of intravenous nitrate therapy for control was regarded as unstable angina. Left ventricular ejection fraction was obtained in all patients by planimetry of a left ventriculogram.
Death within the same hospital admission, regardless of the cause, was defined as in-hospital mortality. Cardiac death was regarded as any death due to cardiac causes, including sudden death. Low output syndrome was defined as the need for postoperative inotropic support or an intraaortic balloon pump for more than 30 minutes to maintain systolic blood pressure greater than 90 mm Hg, mean blood pressure greater than 60 mm Hg, or the cardiac index greater than 2.2 L/(min · m2), despite sufficient volume substitution. Extubation criteria were hemodynamic stability, absence of surgical bleeding, fully rewarming, consciousness and optimal blood gases with fraction of inspired oxygen of 0.3 or less.
Cerebrovascular disease (CVD) was regarded as any transient ischemia accident, reversible ischemic neurological deficit or stroke. Sepsis was defined by at least two of the following clinical criteria: body temperature of less than 36°C or more than 38°C, heart rate exceeding 90 beats/min, respiratory rate exceeding 20 breaths/min, PCO 2 of less than 32 mm Hg, or white blood cell count exceeding 12,000/µL or less than 4000/µL.
Biochemistry
Blood samples were collected preoperatively on the morning of the operation. In 122 patients, CRP samples were available also on the day of operation and postoperative days 1, 2, and 4. High-sensitivity CRP was measured by immunoassay technique (Flex reagent cartridge, Dade Behring Inc, Newark, DE). The sensitivity is 0.05 mg/dL, which represents the lowest concentration of CRP that can be distinguished from zero.
Follow-up
After discharge home, follow-up was accomplished by periodic evaluation of patients at our institution or by contacting the patients’ general practitioners by telephone. Maximum follow-up was 3 years (median, 1.8 ± 1.5 years) and was 92.6% complete (523/565).
Statistical Analyses
Categoric variables were presented as absolute numbers and percentages and were compared with the 
2 test. All continuous variables were tested for normal distribution by evaluation of histograms and the Shapiro-Wilk test. If they were normally distributed (European System for Cardiac Operative Risk Evaluation [EuroSCORE], number of grafts, CPB duration, clamp duration, blood loss), they were expressed as mean ± standard deviation and compared by t test. If not normally distributed (CRP, mechanical ventilation, intensive care unit stay, overall stay, blood units, overall blood products) they were presented as median ± interquartile range (IQR) and were compared with the Mann-Whitney U test. With the necessity to compare the overall magnitude of postoperative CRP release, the linear trapezoidal method was used to evaluate areas under the curve (AUC) of CRP.
Kaplan-Meier curves and log-rank test were used to compare the survival during the follow-up between the two groups. Stepwise logistic regression was used to evaluate the risk-adjusted effect of preoperative CRP 
0.5 mg/dL on short-term outcomes (Appendix Table 1). Cox model was employed to perform risk-adjusted evaluation on long-term outcomes (Appendix Table 2). Preoperative and intraoperative variables that reached significance of 0.2 or less at the first step of multivariable analyses were introduced in the second step (see Appendix Table 1). Statistical analyses were performed using the Stat-View 5.0 statistical software package (SAS Institute Inc, Cary, NC), and NCSS 2001 (Number Cruncher Statistical System, Kaysville, UT).
|
|
|
Results |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
|
Table 2 summarizes short-term results. In-hospital mortality was 3.4% in the LIS group and 8.2% in the HIS group (p = 0.01). Patients in the HIS group had a significantly higher incidence of death caused by sepsis. The incidence of overall postoperative infections was 9.7% (58/597). There were 36 sternal wound infections (6%), only 5 (0.8%) had mediastinitis, the other 31 had superficial infections. Incidence of sepsis was 1.8% (11/597). Of the remaining 11 patients with postoperative infections, 7 had pneumonia and 4 had urinary tract infections. Therefore, according to the criteria for evaluation of postoperative infections recently published by Society of Thoracic Surgeons National Cardiac Database [21], our incidence of major infections was 2.7% (16/597). A preoperative CRP of 0.5 mg/dL or more was associated with significantly higher incidence of infections, especially regarding the sternal wound and sepsis (Table 2).
|
Mid-Term Results
During the follow-up, patients in the LIS group had better survival (91.9% ± 2.5% versus 88.5% ± 2.9%, p = 0.05) (Fig 2) and a better freedom from hospitalization for cardiac events (86.5% ± 3.2% versus 73.6% ± 6%, p = 0.003) compared with the HIS patients. Causes of death in the HIS group were sudden death in 8, acute myocardial infarction in 4, heart failure in 2, stroke in 2, reoperation in 1, and bowel occlusion in 1. Causes of death in LIS group were sudden death in 4, heart failure in 3, acute myocardial infarction in 2, reoperation in 2, neoplasia in 2, and diabetic complications in 1.
|
Cox analysis (see Appendix Table 2) revealed that CRP of 0.5 mg/dL or more was an independent risk factor for mortality during follow-up (OR, 1.93; CI, 1.01 to 4.15; p = 0.05) together with EuroSCORE (OR, 1.25; CI, 1.04 to 1.49; p = 0.01). Moreover, CRP of 0.5 mg/dL or more was an independent risk factor for hospitalization caused by cardiac events (OR 1.82; CI, 1.01 to 3.28; p = 0.05) together with chronic obstructive pulmonary disease (OR 1.79; CI, 0.96 to 3.34; p = 0.06).
|
Comment |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
Our study provides data from a large group of patients in which preoperative inflammatory status was known and offers a different view of the issue. It has been often hypothesized, but not yet clearly demonstrated, that the level of postoperative inflammatory reaction can influence the development of serious complications such as myocardial dysfunction, respiratory failure, renal and neurologic dysfunction, bleeding disorders, and ultimately, multiple organ failure [22].
In our patient population, a preoperative CRP level of 0.5 mg/dL or more was an independent risk factor for in-hospital death. In our analysis of postoperative CRP values of 122 patients, we observed that a more intense inflammatory reaction developed after surgery in patients with an activated inflammatory status. Therefore, preoperative inflammatory status could be considered a risk factor for postoperative inflammatory reaction, magnifying the effects of the intraoperative inflammatory stressors of CPB, ischemia, endotoxemia, and surgical trauma.
A previous prospective study by Gaudino and colleagues [20] of 113 patients undergoing on-pump CABG demonstrated that patients with preoperative CRP values exceeding 0.5 mg/dL had a significantly higher peak of CRP on postoperative day 2, but this study failed to evidence any clinical connection between preoperative inflammatory status and short-term results.
Should CRP be considered as valid as other markers to evaluate inflammation in the context of cardiac surgery? Although specific mechanisms of action have been described for each inflammatory molecule, they are all involved in the same inflammatory process. In fact, a positive and significant correlation between CRP and proinflammatory cytokines has been demonstrated after cardiac operations [20] and in patients with stable or unstable angina [23]. The perioperative evaluation of other inflammatory markers as proinflammatory interleukins or tumor necrosis factor-
on large series of surgical patients would be difficult, expensive, and probably less useful for clinical evaluation of risk.
After decades spent to analyze, in different trials with different aims, the release of several inflammatory markers after cardiac surgery, it is probably time to standardize the evaluation of the inflammatory state before and after cardiac operations. CRP may play such role. The ability of preoperative CRP to predict short-term outcome after cardiac surgery has already been recognized. Biancari and colleagues [17] reported that a preoperative CRP level of 1 mg/dL or more was an independent risk factor for postoperative mortality in 764 CABG patients. Their analysis found that cardiac complications were the main cause of the operative deaths.
Our study showed that a preoperative CRP level of 0.5 mg/dL or more was an independent predictor of overall infections and sternal wound infections. Furthermore, patients with this level of CRP experienced a higher incidence of sepsis, death caused by sepsis, and a prolonged time of mechanical ventilation. Similarly, Boeken and colleagues [18], studying 100 patients who had undergone coronary and valve surgery, showed that a preoperative CRP level exceeding 0.5 mg/dL was associated with a higher incidence of septic complications, need of inotropic support, and a significantly longer duration of mechanical ventilation and intensive care unit stay. Accordingly, Fransen and colleagues [19] evidenced that CRP exceeding 0.8 mg/dL, together with diabetes mellitus and CPB time of more than 112 minutes, were independent risk factors for postoperative infections.
A possible explanation for the higher rate of postoperative infective complications in patients with high preoperative inflammatory status lies in the observation that bacterial growth is increased in presence of elevated concentrations of proinflammatory mediators [24]. In their study of patients with acute respiratory distress syndrome, Meduri and colleagues [25] showed that ventilator-associated pneumonia and consequent mortality is related to the magnitude and duration of the host inflammatory response. The excessive and nonspecific inflammatory response observed in the HIS group may favor, similar to patients with acute respiratory distress syndrome, perioperative infections. Future studies should evaluate how the immune system reacts to an infectious insult depending on previous activation or nonactivation of the inflammatory system. To improve the operative outcome, strategies able to decrease the perioperative inflammatory reaction might be adopted, particularly in patients with an activated preoperative inflammatory state.
The interpretation of high preoperative CRP values may be difficult: it can be due to an underlying occult infection or to the extension and severity of cardiac pathology (eg, unstable angina, recent myocardial, congestive heart failure). White blood cell evaluation cannot always exclude the presence of occult infection, but so far, we do not have other suggestions to improve the preoperative evaluation. We believe, however, that in most of the HIS patients, high preoperative CRP values were the consequence of their cardiac pathology rather than infection; in this case, surgery is often needed without delay.
The original finding of our study is that patients with an activated inflammatory status before cardiac surgery had a worse outcome in the mid-term. A CRP value of 0.5 mg/dL or more was an independent risk factor for mortality and hospitalization for cardiac events during follow-up. Several large studies assessed the power of CRP to predict future cardiovascular events in a healthy population [1–4]. In particular, Ridker and colleagues [2], studying 27,939 healthy women and dividing them on the basis of nine clinically set CRP cut-points, highlighted that CRP exceeding 0.5 mg/dL increased by 2.3-fold the risk of a first cardiovascular event over an 8-year follow-up period.
In patients undergoing myocardial revascularization through PCI, prospective studies evidenced that preprocedural CRP affects survival and restenosis rate [7–9]. Dibra and colleagues [7] observed in 1152 patients who underwent coronary artery stenting that preprocedural CRP exceeding 0.5 mg/dL increased by almost twofold the risk of death or myocardial infarction at 1 year. It was quite surprising that similar results occurred in our population in whom surgical correction of cardiac disease had been performed.
Our postoperative CRP evaluations showed that after a peak reached on postoperative day 2, inflammation tended to decrease on postoperative day 4, but unfortunately, we are not able to provide data describing the inflammatory state during the follow-up. It is possible that the HIS patients, despite surgical correction of their disease, continue to have a highly activated inflammatory state. The relationships between inflammatory cascade and atherosclerosis are well known: initiation, growth, and complication of an atherosclerotic plaque seems to be the inflammatory response of the endothelium to the injuries produced by risk factors as hypertension, smoking, diabetes mellitus, and hypercholesterolemia.
Initially suggested to be a biomarker, CRP now appears to be a mediator of atherogenesis [26], able to upregulate nuclear factor-
B, a key nuclear factor that facilitates the transcription of numerous proatherosclerotic genes [27]. CRP activates endothelial cells and increases plasminogen activator inhibitor-1 (PAI-1) expression and activity in human endothelial cells [28]. PAI-1 activity measured on the first postoperative day was significantly higher in patients who were later found to have vein graft occlusion at angiographic evaluation 3 months after CABG [29].
Although numerous investigators have documented profound short and mid-term coagulative-fibrinolytic and inflammatory alterations after cardiac surgery, no clinical studies to date have prospectively evaluated, in an appropriate number of patients, the value of prothrombotic and proinflammatory markers in predicting early graft occlusion and mid-term clinical outcome. Patients with a highly activated preoperative inflammatory state might preserve such status after the operation, with consequent progression of the atherosclerotic disease and a higher risk of early graft occlusion.
It should be noted that large retrospective studies have shown that preoperative statin treatment is associated with decreased morbidity and mortality after cardiac operations [30]. So far, no clinical studies have evaluated the effects of statin administration during the postoperative period on the incidence of early graft occlusion and on the short-term and mid-term proinflammatory and prothrombotic changes induced by CABG surgery. Statin treatment in patients with a CRP value of 0.6 mg/dL or more undergoing PCI reduces the incidence of death due to cardiac causes, myocardial infarction, and repeat intervention of the stented vessel [9]. The effects of statin treatment on the prothrombotic state following cardiac surgery might be promising.
Patients with an activated inflammatory status who underwent cardiac operations showed an increased inflammatory reaction to cardiac surgery. A preoperative CRP level of 0.5 mg/dL or more significantly affects short-term and mid-term outcome in patients undergoing cardiac surgery. Limits of the study are the lack of a prospective design and the absence of data on the inflammatory state and statin therapy during follow-up. Future trials are needed to ascertain a mid-term protective role of antiinflammatory agents in patients with an elevated inflammatory status who undergo cardiac surgery.
|
References |
|---|
|
TopAbstractIntroductionMaterial and MethodsResultsCommentReferences |
|---|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| ANN THORAC SURG | ASIAN CARDIOVASC THORAC ANN | EUR J CARDIOTHORAC SURG |
| J THORAC CARDIOVASC SURG | ICVTS | ALL CTSNet JOURNALS |
