Preoperative C-reactive protein level and outcome following coronary surgery

Abstract

Objective: It has been suggested that a preoperative level of C-reactive protein (CRP)>5 mg/l is predictive of postoperative complications in cardiac surgery patients. Material and methods: Among 113 primary isolated coronary artery bypass patients, CRP was determined preoperatively and fibrinogen, interleukine 6, plasminogen activator inhibitor-1, prothrombin time, activated partial thromboplastin time, platelets and white blood cells count measured before surgery, 24, 48 and 72 h thereafter and at hospital discharge. The clinical course of all cases was prospectively recorded. Data were then analysed according to the preoperative CRP level by dividing the patients into two groups (CRP>5 mg/l or CRP<5 mg/l). Results: The in-hospital results were similar between the two patients groups. Even the postoperative haematic inflammatory markers did not significantly differ according to the preoperative CRP level. Conclusion: In this prospective study, a preoperative level of CRP>5 mg/l did not predict in-hospital postoperative complications nor influence the extent of the inflammatory activation in primary isolated coronary bypass patients.

1 Introduction

C-reactive protein (CRP) is a classical infection and acute phase reaction indicator. Controversies exist on its role for the preoperative identification of patients at high perioperative risk after cardio-pulmonary bypass (CPB) procedures, with different authors suggesting and denying in turn a correlation between preoperative CRP level and postoperative clinical course [1–3]. The present investigation was intended to evaluate the correlation between the pre- and postoperative levels of CRP and other inflammatory and fibrinolytic markers and the postoperative outcome in a large prospective trial.

2 Patients and methods

2.1 Patients population

The patients included in this report represent the all study population of the Fibrinolisi e Infiammazione nella Fase Acuta (FIFA) study. This study was a prospective randomized investigation on the inflammatory and fibrinolytic reaction to cardiopulmonary bypass whose results have been in part already reported [4,5]. From January 1998 to May 1999, all patients scheduled to undergo isolated elective coronary revascularization procedures at our Institution were screened for inclusion. Exclusion criteria were: associated cardiac or non-cardiac surgical procedures, age >80 years, single-vessel disease, emergent or urgent revascularization, left ventricular ejection fraction <0.30, carotid artery disease, previous cerebrovascular accident, chronic dialysis, hepatic failure, respiratory or renal insufficiency, hemorrhagic conditions, active infection and chronic anti-inflammatory therapy. To rule out possible confounding effects of circadian variability, only patients operated between 8 and 10 a.m. were considered. Overall 113 cases were included.

2.2 Operative technique

The anaesthetic procedures were standardized for all patients: they received their medication until the operative day and were premedicated with diazepam (0.15 mg/kg), morphine (0.15 mg/kg) and scopolamine (0.01 mg/kg). After induction with sodium thiopental (2–3 mg/kg) balanced anaesthesia was performed with isoflurane, fentanyl and boluses of midazolam and propofol as needed to maintain hypnosis.

Muscle relaxation was ensured by administration of pancuronium bromide (0.1 mg/kg). Anticoagulation was obtained with heparin 300 IU/kg and an activated clotting time >480 s. maintained. Heparin neutralization was achieved with prothamine hydrochloride 1.3 mg/mg heparin. All surgical procedures were performed in standard fashion by the same surgical team through median sternotomy, and using CPB. During CPB, the nasopharingeal temperature was kept at 37°C in 55 cases and at 26°C in 58. Myocardial protection was always accomplished by anterograde isothermic intermittent blood cardioplegia.

2.3 Postoperative evaluation

The clinical course of all patients, including major and minor postoperative complications, mean duration of mechanical ventilation and stay in the intensive care unit and in hospital were prospectively recorded. CRP, fibrinogen, interleukine 6 (IL-6), plasminogen activator inhibitor-1, prothrombin time, activated partial thromboplastin time, platelets and white blood cells counts were measured the day before surgery, 24, 48 and 72 h thereafter and at hospital discharge.

2.4 Measure and analysis method

All the blood samples were collected immediately before surgical operation, 24, 48 and 72 h after and at hospital discharge.

Plasmatic concentration of all the molecules considered were obtained as follows:

• CRP was determined by using nephelometric method (APS, Beckman, Palo Alto, USA).

• Fibrinogen was determined by using Clauss method (Dade, Miami, FL; MLA 1600)

• IL-6 was determined by using Elisa IL-6 system (Human Biotrak Elisa System, Amersham Pharmacia Biotech, Monza, Italy).

• Plasminogen activator inhibitor-1 (PAI-1) antigen levels were determined by double antibody sandwich enzyme-linked immunosorbent assay (Instrumentation Laboratory SpA, Milano, Italy).

The remaining parameters (international normalized ratio (INR), activated partial thromboplastin time (aPTT), platelets and white blood cells counts) were measured in the hospital central laboratory and used for the clinical management before and after surgery.

2.5 Statistical analysis

Qualitative data were analysed by chi-square test. Quantitative data were expressed as mean±standard deviation (SD) of the mean and analysed by t test for independent samples. Quantitative repeated measurements data were analysed by two factor (time and treatment) analysis of variance (ANOVA) for repeated measures; post-hoc Newman Keuls test was performed if P<0.05. A P value <0.05 was considered significant.

With reference to preoperative levels of CRP, the population was divided into two groups: CRP>5 mg/l or CRP<5 mg/l (high and low group, respectively).

3 Results

3.1 Overall population

There were two in-hospital deaths (1.7%), respectively, from massive pulmonary embolism and myocardial infarction. Postoperatively, five cases of myocardial infarction, one of renal insufficiency, one sepsis and five revisions for bleeding were recorded. Mean postoperative stay in the intensive care and in-hospital were 2.0±2.7 and 6.1±3.2 days, respectively.

The time course of the dosed inflammatory and fibrinolytic markers in the overall population is summarized in Figs. 1–4 . As shown in Fig. 1, fibrinogen decreased in the first 24 h after surgery and then started rising until discharge (acting as a late acute phase protein), whereas PTT and INR had a similar pattern with a slight increase immediately after surgery followed by a substantial stability during the rest of the hospital stay. Fig. 2 depicts the PAI course: PAI activity peaked in the first postoperative hours and returned to baseline on the third postoperative day while PAI antigen showed two separate peaks, one 48 h after surgery and the second immediately before discharge. Interleukin-6 peaked in the first postoperative day and then gradually returned to baseline whereas CRP started rising from the second postoperative day and peaked 72 h after surgery (Fig. 3). Finally haematological values summarized in Fig. 4 showed that white blood cells, neutrophils and lymphocytes counts peaked immediately after the operation and then returned to baseline at the time of hospital discharge, monocytes had a biphasic time course with peaks 24 h after surgery and at discharge and platelets were significantly reduced after surgery and returned to baseline value only at the time of hospital discharge.

As reported elsewhere [4], no effect of CPB temperature on any of the measured variables at any time point could be demonstrated.

3.2 Group high vs. group low

3.2.1 Clinical data

Group high consisted of 31 patients vs. 82 of group low; the mean age was 61±8 vs. 62±10 years old. Concerning the preoperative risk factors 22 cases in group high had hypertension vs. 49 in group low, 5 vs. 18 were diabetics and 5 vs. 25 hypercholesterolemic. Eighteen patients in group high had a history of myocardial infarction and two suffered of heart failure vs. 39 and 3, respectively, in the group low. Left main coronary artery disease was present in five patients in group high vs. nine in group low. None of these differences was statistically significant.

Intraoperatively, CPB and cross-clamp time were also similar for the two groups (respectively, 75.37±20.6 vs. 74.9±18.8 and 62.5±21.7 vs. 61.8±16.8 min; P 0.92 and 0.85) as was the number of graft per patient (3.1±0.7 vs. 3.2±0.8; P=0.40).

The in-hospital clinical results were similar in both CRP groups. Particularly there were no differences regarding incidence of postoperative renal failure (1 vs. 0), assisted ventilation >24 h (6 vs. 1), inotropic drugs administration >24 h (0 vs. 0), need for blood transfusions (20 vs. 8), and need for surgical revision for bleeding (3 vs. 2). Two in-hospital deaths (1 vs. 1) occurred due to massive pulmonary embolism (on the third postoperative day) and myocardial infarction (8 days after surgery). Postoperatively, five patients experienced myocardial infarction (1 vs. 4), one patient had a septic syndrome (0 vs. 1) and two cases had to be reoperated for clinical or instrumental evidence of graft malfunction (0 vs.2). None of these differences was statistically significant.

Mean stay in intensive care and in-hospital after surgery in group high and low were 1.9±2.6 vs. 1.9 ±1.8 and 6.1±3.2 vs. 5.4±1.9 days, respectively (P=0.9 and 0.3).

3.2.2 Inflammatory markers trend analysis

The postoperative haematic inflammatory markers average in the two groups concentration trend is showed in Figs. 7 and 8. Average level of postoperative CRP in both CRP groups increased in the postoperative period with a peak concentration at 48 h for the high group; low group CRP levels grow up significantly slower than the group high and reached a significantly lower peak 24 h later. Elevated CRP levels were maintained until hospital discharge in both groups. The average values of IL-6, white blood cells, lymphocytes, neutrophils and monocytes showed small and non-significant differences between the two groups (see Figs. 7 and 8).

3.2.3 Fibrinolysis markers trend analysis

There were no differences in INR, PTT and PAI between the two CRP groups. However, patients with preoperative CRP>5 mg/l showed a significantly superior levels of platelets and fibrinogen all throughout the postoperative sampling period (see Figs. 5 and 6 ).

4 Discussion

With almost 800,000 operations worldwide per year coronary artery bypass grafting (CABG) is one of the most frequent surgical procedure actually performed. The technical and technological improvements occurred in the last decade have rendered this operation more and more safe, reducing to acceptable levels the surgical risk even in very old and sick cases. Despite that, according to the data of the Society of Thoracic Surgeons cardiac surgery database, about 10% of patients submitted to CABG experience significant morbidity and more than 30% of them face a major postoperative complication.

In view of these considerations, the possibility of predicting the hospital course of a patient by using a simple, relatively inexpensive and readily available markers such as the CRP would be of obvious interest.

Boralessa and associates in 1986 first suggested a possible predictive role of CRP in cardiac surgery patients: in a series of 25 patients undergoing coronary and valvular surgery these authors found that all cases with normal preoperative CRP levels recovered uneventfully, whereas the great majority of those with raised CRP values before surgery and persistent elevation after the operation experienced major postoperative complications [1]. These data were contradicted the following year by the observations of Kress et al. who, in a large cohort of 80 cardiac surgery cases could not confirm the predictive role of CRP with regard to the postoperative outcome [2].

However, more recently Boeken et al. retrospectively reviewed the files of a group of 50 patients undergoing coronary and valvular procedures with a preoperative CRP level >5 mg/l and compared their results with those of a cohort of 50 matched cases with a normal preoperative CRP value, finding a superior incidence of postoperative complications among the farmers [3]. As microbiological tests were negative, these authors attributed the more complicated postoperative course to the systemic inflammatory reaction to CPB and underscored the important prognostic value of CRP (suggesting even deferral of the operation in case of elevated preoperative level).

In our series no correlation could be established between the preoperative CRP value and the postoperative clinical course: both the type and incidence of major and minor postoperative complications were similar in the high and low groups. Although, as already described, the peak time and value and the pattern of normalization of CRP were different in the two series, other subtle markers of inflammation (such as IL-6 and the white blood cells count) were similar in the two patients groups (see Figs. 7 and 8 ).

These observations seems then to deny a superior intensity of the inflammatory reaction to CPB in the high group and to minimize the potential predictive value of CRP.

The different results reported by Boralessa and Boeken [1,3], are probably related to the retrospective methodology used by the previous authors or the limited statistical power (as this is by far the largest study on this issue). Moreover in none of the previous series the entity of the systemic reaction to CPB was determined using sensible laboratory markers.

Finally, the evidence of a different postoperative pattern in platelets and fibrinogen levels (with higher values in the high series), although without any clinical correlation, might suggest a greater thrombotic attitude in patients with preoperative CRP>5 mg/l and should act as a stimulus for further investigation during the follow-up.

In conclusion, our data furnish substantial evidence that elevated preoperative CRP cannot be considered a marker of increased surgical risk and CRP level should not influence the decision process before cardiac surgery procedures.

References