INTRODUCTION

Each year, approximately 795,000 people experience a new or recurrent stroke in the United States of America. Approximately 610,000 of these are first attacks. Of all strokes, 87% are ischemic^[1]. In Brazil, stroke is one of the leading causes of death. Despite encouraging data regarding age-standardized declining stroke incidence and mortality in Brazil and globally, the aging population and accumulation risk factors contribute to an increased lifetime risk of stroke^[2].

Stroke remains a devastating complication after cardiac surgical procedures with substantial functional and economic impacts^[3]. Despite all efforts, the incidence of postoperative stroke remains as high as 6%^[4]. Even when patients survive to discharge, they have increased hospital length of stay and are more likely to need to be discharged to a nursing home or a rehabilitation center. Thirty-day mortality is also increased in these patients^[5]. There are many attributed causes of postoperative stroke. Carotid artery stenosis (CAS) and aortic plaque load are predominantly responsible. However, cardiac emboli, cerebral vasospasm, particulate emboli from cardiopulmonary bypass (CPB), transient hypercoagulable state, and hypoperfusion have also been implicated. The presence of multiple risk factors in the majority of patients makes understanding the exact etiology challenging^[6].

Previous studies have attempted to identify predictors of stroke after coronary artery bypass grafting (CABG) with or without combined valvular procedures. These factors seem to be related not only to clinical characteristics (e.g., advanced age, heart failure, diabetes mellitus) but also to the process of care, such as prolonged CPB^[7-10]. The results are so far conflicting or incomplete mainly due to a low number of interventions or a low number of variables^[11]. Additionally, Brazil and other Latin-American countries lack contemporary statistics regarding stroke risk after cardiac surgery.

Therefore, the present study aimed to investigate the incidence and predictors of perioperative and seven-day postoperative stroke in a retrospective database from a university hospital.

METHODS

RESULTS

The population consisted of 678 patients who underwent cardiac surgery between July 1, 2011, and December 31, 2018. In the overall sample, mean age was 61.4±10.7 years, and men comprised 469 (69.2%) patients. Medical comorbidities include hypertension in 556 (82%), type 2 diabetes mellitus in 240 (35.4%), and coronary artery disease in 522 (77%) patients.

Surgical procedures were distributed as follows: CABG in 461 (68%), aortic valvular replacement in 85 (12.5%), and mitral valve replacement in 44 (6.5%) patients. Table 1 describes the clinical characteristics of the sample and the surgical procedure.

Table 1 - Epidemiologic characteristics of the population.

Overall mortality in the population was 47 (7.1%) cases. The main cause of death in the population was cardiogenic shock (3.4%), followed by sepsis (2.1%). Nineteen (2.8%) patients required surgical reintervention within 30 days. The main reasons for reinterventions were bleeding (1.1%), mediastinitis (0.4%), and cardiac tamponade (0.4%). The mean length of hospital stay was 16 days (interquartile range 11-25), with intensive care unit (ICU) stay of five days (interquartile range 4-7). Thirteen patients (2%) were readmitted to the ICU.

The incidence of postoperative infection was 209 (31.1%) cases, with the following distribution: pulmonary infection in 116 (17.1%) and surgical wound infection in 49 (7.2%) patients. Postoperative stroke occurred in 24 patients (3.5%), 23 (3.3%) were ischemic, and 21 (3.0%) were diagnosed in the first 72 hours after surgical procedure. Two individuals had a presumptive diagnosis of stroke after neurological evaluation but were excluded from the sample because they did not have a complementary imaging exam. The other outcomes are shown in Table 2.

Table 2 - Postoperative outcomes.

The univariate predictors of seven-day ischemic stroke were determined in Table 3. After multivariate analysis, the following factors were significantly associated with stroke: previous stroke/transient ischemic attack (TIA) (relative risk [RR]=2.75; 95% CI, 1.11-6.82; P=0.029), carotid artery disease (RR=4.0; 95% CI, 1.43-11.0; P=0.008), previous AF (RR=3.26; 95% CI, 1.31-8.1; P=0.011), and postoperative platelets > 200,000/mm³ (RR=2.26; 95% CI, 1.01-5.1; P=0.048) (Table 4).

Table 3 - Risk factors for seven-day stroke in univariate analysis.

Table 4 - Risk factors for seven-day stroke in multivariate analysis.

DISCUSSION

Neurologic complications after cardiac surgical procedures remain relatively common despite improvements in anesthetic and surgical techniques, as well as in perioperative monitoring and management^[14]. Previous studies have demonstrated that neurologic deficits occur in as many as 6% of patients undergoing cardiac surgery^[5,9,15,16]. However, in a contemporary metanalysis, Gaudino et al.^[14] described a pooled event rate of perioperative stroke of 2.03%, with an incidence of ischemic stroke of 88%. The incidence of clinically silent cerebral infarction though is higher (up to 18%), particularly after aortic valve replacement surgery^[4]. The occurrence of early stroke is associated with a significant increase in mortality as well as much higher increase in the risk of late death, suggesting that addressing this potential complication can significantly improve the outcomes of cardiac surgery^[17].

In this single-center study, we determined predictors of early (perioperative and seven-day postoperative) stroke with CABG, heart valve surgery alone, or heart valve surgery combined with CABG. We observed a stroke rate of 3.5% (24 patients), with 3.3% (23 patients) suffering an ischemic event. Additionally, 3.0% of the strokes (21 cases) were diagnosed in the first 72 hours after the surgical procedure. These numbers are high compared to the contemporary series of 2.3% of perioperative ischemic strokes^[14]. In the last decades, significant efforts have been aimed at perioperative stroke reduction, including minimizing aortic manipulation, eliminating CPB, and using preoperative CT scan of the ascending aorta or duplex scanning of the carotid arteries as well as epiaortic ultrasound^[18-20]. We reported a higher number of such complication, and the reasons may be related to the lack of these strategies in our service, which suggest that early stroke is probably linked to intraoperative events, reflecting the technical or surgical nature of its etiology^[14,17,21]. We also must acknowledge that the present series represents data of a single cardiology center with only two surgeons available and low annual surgery rate. As previously reported, increased surgeon and hospital volume significantly improves outcomes in cardiovascular surgery, although the benefit is less pronounced than in other specialties^[22].

After multivariate analysis, we determined the following independent factors associated with stroke: previous stroke or TIA^[23], carotid artery disease, previous AF, and early postoperative platelet counts > 200,000/mm³. In the present series, we demonstrated that a history of previous stroke or TIA was an independent risk factor for the development of postoperative stroke, increasing the risk of the event by 2.75. Similar findings were reported by Hogue et al.^[16]. Additionally, the documentation of significant carotid artery disease, with stenosis > 50% in internal carotid arteries, was associated with four times the risk of stroke. In a Brazilian series by Costa et al.^[24], the rate of ischemic stroke in patients with carotid stenosis ≥ 50% was 10%, while in patients with carotid stenosis ≥ 70% this index reached 16.6%. It is suggested that up to 30% of early postoperative strokes complicating CABG are caused by a hemodynamically significant CAS; however, some evidence disputes this presumption. Contemporary data suggest that the single most significant marker of an adverse cerebral outcome after CABG is the identification of atherosclerotic plaques in the ascending aorta, suggesting carotid lesions are more a risk marker than the etiology of stroke. The high prevalence of aortic atherosclerosis in patients with CAS may help to explain the higher stroke rate in patients with lesions outside the territory of their carotid disease. This also creates great doubt about the benefits of prophylactic treatment of carotid lesions in patients without neurological symptoms^[25].

Regarding arrhythmia, the presence of AF was an independent predictor of stroke in our study, with an RR of 3.26. The embolic mechanism associated with clots in the left atrium is the most probable explanation for this association, particularly in patients with unsatisfactory control of anticoagulation or intraoperative surgical manipulation or spontaneous recovery of sinus rhythm postoperatively^[26]. Pierik et al.^[27] performed a secondary analysis from a randomized trial and observed that patients with POAF within the first seven days after cardiac surgery had a three-fold increased risk for thromboembolic stroke during hospital admission based on the difference in the incidence of POAF between patients with and without thromboembolic stroke. In a meta-analysis published by Kaw et al.^[28] and Lin et al.^[29], the development of new-onset POAF after cardiac and non-cardiac surgery was associated with increased shortand long-term stroke and mortality. Several series have confirmed a higher long-term mortality after the development of POAF in cardiac surgery^[30-32]. There are published guidelines to support the prevention of AF in cardiac surgical patients, but there is insufficient evidence to recommend rhythm control for stroke prevention once fibrillation has occurred^[33].

Finally, our study found postoperative platelet counts > 200,000/mm³ as an independent risk factor for stroke. This finding is somehow surprising and opposite to previous evidence. Karhausen et al.^[34] demonstrated in a single-center study that moderate to severe postoperative thrombocytopenia was an independent risk factor for the development of postoperative stroke. Griffin et al.^[35] showed that postoperative thrombocytopenia is independently associated with postoperative mortality, acute kidney injury, infection, stroke, and prolonged ICU and hospital lengths of stay. The mechanism for developing thrombocytopenia in these patients remains unclear, but the association of postoperative platelet counts with stroke after CABG surgery suggests that the reduction in platelet numbers likely occurs in a context of increased platelet reactivity^[34].

CONCLUSION

We demonstrated that a history of previous stroke or TIA, carotid artery disease, previous AF, and platelet counts > 200,000/mm³ showed as independent risk factors for the development of postoperative stroke.