β Blockers are widely used to improve the postoperative cardiac outcome in patients with coronary artery disease scheduled for noncardiac surgery. However, recently serious concerns regarding the safety of perioperative β blockers have emerged. To assess the incidence, risk factors, and β-blocker use associated with postoperative stroke in the Dutch Echocardiographic Cardiac Risk Evaluation Applying Stress Echocardiography (DECREASE) trials, we evaluated all 3,884 patients of the DECREASE trials for postoperative stroke. All cardiac risk factors and medication use were assessed. The incidence of stroke within 30 days after surgery was recorded. The incidence of postoperative stroke in the DECREASE trials was 0.46% (18 of 3,884). For the β-blocker users, the incidence was 0.5%. All the strokes had an ischemic origin. A history of stroke was associated with a greater incidence of postoperative stroke (odds ratio [OR] 3.79, 95% confidence interval [CI] 1.2 to 11.6). Statins and anticoagulants were not associated with postoperative stroke (OR 0.85, 95% CI 0.3 to 2.4; and OR 1.27, 95% CI 0.4 to 4.6, respectively). No association with bisoprolol therapy was found (OR 1.16, 95% CI 0.4 to 3.4). In conclusion, with a low-dose bisoprolol regimen started ≥30 days before surgery, no association was observed between β-blocker use and postoperative stroke.
β Blockers are prescribed in the perioperative period to reduce myocardial oxygen consumption by decreasing sympathetic tone and myocardial contractility. Several randomized clinical trials have shown beneficial effects of initiating β blockers to patients at risk of cardiovascular complications, but other trials have failed to show any significant improvement in outcome. To resolve these inconsistencies, a large multinational randomized controlled trial was started in 2002: the PeriOperative ischemic Evaluation (POISE) trial.
The POISE trial randomly assigned 8,351 patients with, or at risk of, atherosclerotic disease who were undergoing noncardiac surgery to receive extended-release metoprolol succinate (n = 4,174) or placebo (n = 4,177). In the POISE trial, high doses of metoprolol, ≤400 mg, were administered in 4 doses starting 2 to 4 hours before surgery and continued for 30 days. The 30-day results showed a significant reduction in cardiac events. However, this came at the cost of a significant increase in the incidence of total mortality and stroke. The high incidence of stroke found in the POISE trial made us question the liberal use of β blockers during the perioperative period.
In the Dutch Echocardiographic Cardiac Risk Evaluation Applying Stress Echocardiography (DECREASE) trials, a low-dose β-blocker regimen was used. No relation was found between stroke and β-blocker use. However, that experience was based on individual DECREASE studies with associated weak statistical power. Therefore, we combined the results of the DECREASE I, II, and IV studies to evaluate the effect of low-dose β-blocker use in the perioperative period on postoperative stroke and to identify the risk factors for postoperative stroke.
Methods
We performed a pooled analysis of all 3,884 patients, who were enrolled in the DECREASE I (1996 to 1999), DECREASE II (2000 to 2005), and DECREASE IV (2004 to 2008) trials. We did not include the DECREASE III trial, because it was a randomized controlled trial in which vascular surgery patients were randomized for statins or placebo. The DECREASE I trial was a randomized controlled trial that clearly demonstrated the cardioprotective effect of bisoprolol in high-risk patients undergoing major vascular surgery. The aim of the DECREASE II study was to assess the value of cardiac testing according to the American College of Cardiology/American Heart Association guidelines for intermediate-risk patients receiving β-blocker therapy with tight heart rate control scheduled for major vascular surgery. The DECREASE IV trial was a large prospective, randomized trial that sought to assess the effect of different pharmaceutical strategies (β blockers or statins, or both) to prevent perioperative cardiac complications in intermediate-risk patients. The results and study protocols have been previously published in detail.
The following potential risk factors for postoperative stroke were recorded from all the trials and included in the present analysis: age >70 years, angina pectoris, previous myocardial infarction on the basis of history or a finding of pathologic Q waves on the electrocardiogram, compensated congestive heart failure or a history of congestive heart failure, drug therapy for diabetes mellitus, renal dysfunction (serum creatinine >160 μmol/L), and previous stroke or transient ischemic attack. Throughout the DECREASE I, II, and IV studies, the definition of the several risk factors was not changed. Also, the type and dosage of perioperative β blockers were noted. The β-blocker dose was converted to a percentage of the maximum recommended therapeutic dose (MRTD) according to the Food and Drug Administration’s Center for Drug Evaluation and Research database. The MRTD for atenolol is 3.330 mg/kg body weight/day; for bisoprolol, 0.330 mg/kg body weight/day; for metoprolol, 6.670 mg/kg body weight/day; for carvedilol, 0.417 mg/kg body weight/day; for propranolol, 10.700 mg/kg body weight/day; and for labetalol, 40.700 mg/kg body weight/day.
The end point for the present study was stroke, either a transient ischemic attack or a cerebrovascular accident within 30 days after surgery. Stroke was confirmed by computed tomography according to the criteria stated in the guidelines from the Stroke Council, American Heart Association. As per protocol, all patients suspected to have had an in-hospital stroke were evaluated by computed tomography.
Because the results of the POISE trial were our incentive to perform the present analysis, we also compared the rate of strokes recently reported for the POISE trial. We also compared both studies concerning the possible risk factors for postoperative stroke.
The continuous data are described as the mean ± SD or the median and range. Dichotomous data are described as numbers and percentages. Logistic regression analysis was used to identify the clinical characteristics and medical therapy (e.g., statins, β blockers, antiplatelet agents, oral anticoagulants) associated with postoperative stroke. Interaction terms were included if statistically significant. Odds ratios (ORs) are reported with the corresponding 95% confidence intervals (CIs). For all tests, p <0.05 (2-sided) was considered significant. All statistical analyses were performed using Statistical Package for Social Sciences statistical software (SPSS, Chicago, Illinois).
Results
The baseline clinical characteristics of the patients included in the DECREASE I, II, and IV trials are listed in Table 1 . The key characteristics were as follows: mean age 67 years, 71% men, 16% with diabetes mellitus, 28% with angina pectoris, 26% with previous myocardial infarction, and 11% with a history of cerebrovascular disease. Overall, 37% of the patients were receiving statin therapy; the patients in the DECREASE IV study were randomized for statin therapy. Most patients (73%) underwent vascular surgery. The incidence of postoperative stroke in the DECREASE trials was 0.46% (18 of 3,884). In the DECREASE I, II, and IV trials, 3, 8, and 7 patients experienced stroke, respectively. The baseline characteristics of the patients with stroke are listed in Table 2 . Perioperative β-blocker use was present in 12 (67%) of 18 patients who experienced a postoperative stroke. The average dose of bisoprolol used among the stroke patients was 15% of the MRTD. Among the β-blocker users, the incidence was 0.5% (12 of 2,366). All strokes had an ischemic origin. The median day on which the strokes had occurred was the second postoperative day (interquartile range 1 to 4). A history of stroke was associated with postoperative stroke (OR 3.79, 95% CI 1.2 to 11.6). Statins and anticoagulants were not associated with postoperative stroke (OR 0.85, 95% CI 0.3 to 2.4; and OR 1.27, 95% CI 0.4 to 4.6, respectively). No association with bisoprolol therapy was found (OR 1.16, 95% CI 0.4 to 3.4; Table 3 ).
| Variable | DECREASE I Trial (n = 1,351) | DECREASE II Trial (n = 1,467) | DECREASE IV Trial (n = 1,066) | DECREASE Trials Overall (n = 3,884) | POISE Trial (n = 8,351) |
|---|---|---|---|---|---|
| Age (years) | 67.5 | 67.0 | 66.8 | 67.1 | 69.0 |
| Men | 1,050 (77%) | 1,075 (73%) | 639 (60%) | 2,764 (71%) | 5,293 (63%) |
| Ischemic heart disease | 1,596 (41%) | 3,589 (43%) | |||
| Myocardial infarction | 562 (42%) | 396 (27%) | 54 (5%) | 1,012 (26%) | |
| Angina pectoris | 229 (17%) | 808 (55%) | 55 (5%) | 1,092 (28%) | |
| Heart failure | 96 (7%) | 168 (11%) | 8 (1%) | 272 (7%) | 220 (3%) |
| Diabetes mellitus | 187 (13%) | 320 (21%) | 115 (11%) | 622 (16%) | 2,427 (29%) |
| Cerebrovascular disease | 104 (8%) | 280 (19%) | 46 (4%) | 428 (11%) | 1,263 (15%) |
| Renal dysfunction | 59 (4%) | 124 (8%) | 11 (1%) | 194 (5%) | 401 (5%) |
| Medication use | |||||
| Statins | 285 (21%) | 617 (42%) | 533 (50%) | 1,435 (37%) | 2,677 (32%) |
| Angiotensin-converting enzyme inhibitors | 394 (29%) | 478 (32%) | 96 (9%) | 968 (25%) | 3,717 (44%) |
| Aspirin | 406 (30%) | 664 (45%) | 102 (9%) | 1,174 (30%) | 3,011 (36%) |
| Variable | n = 18 |
|---|---|
| Age (years) | 71.6 |
| Men | 13 (72%) |
| Myocardial infraction | 5 (27%) |
| Angina pectoris | 1 (5%) |
| Heart failure | 0 (0%) |
| Diabetes mellitus | 3 (16%) |
| Cerebrovascular disease | 6 (33%) |
| Renal dysfunction | 1 (5%) |
| Medication use | |
| β Blocker | 12 (67%) |
| Statins | 11 (32%) |
| Angiotensin-converting enzyme inhibitors | 4 (22%) |
| Aspirin | 8 (44%) |
| Surgery type | |
| Colorectal | 1 (5%) |
| Esophagogastric | 3 (16%) |
| Ear/nose/throat | 1 (5%) |
| Urologic | 1 (5%) |
| Orthopedic | 1 (5%) |
| Vascular | 11 (61%) |
| Variable | Odds Ratio | 95% Confidence Interval |
|---|---|---|
| Medical risk factors | ||
| Age (per year increase) | 1.04 | 0.99–1.10 |
| Male gender | 1.41 | 0.45–4.42 |
| Ischemic heart disease | 0.4 | 0.13–1.21 |
| Cerebrovascular disease | 3.79 | 1.24–11.6 |
| Diabetes mellitus | 0.88 | 0.31–3.93 |
| Medication use | ||
| β Blockers | 1.16 | 0.40–3.35 |
| Statins | 0.85 | 0.30–2.40 |
| Antiplatelet therapy | 1.75 | 0.60–5.15 |
| Oral anticoagulants | 1.27 | 0.35–4.64 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
