Misconduct May Take Bloom Off β-Blocker Rx
by Karen Blum
The debate over using β-blockers perioperatively for non-cardiac surgery patients continues with a new meta-analysis from British researchers indicating that starting the drugs specifically for surgery can increase a patient’s risk for death by 27%.Reporting in Heart (doi:10.1136/heart-jnl-2013-304262 [Epub ahead of print]), investigators from Imperial College, in London, reviewed nine studies of perioperative β-blocker use in 10,529 patients, and found that β-blockers caused a “statistically and clinically significant increase in mortality.” Analyzing six of the studies, the team found that β-blockers reduced the risk for non-fatal myocardial infarction (MI) (relative risk [RR], 0.73; P=0.001) but increased the risk for stroke (RR, 1.73; P=0.05) and hypotension (RR, 1.51; P<0.00001).
Meanwhile, U.S. and European cardiology guidelines released in 2009 recommend this therapy, and the study authors say it’s time for an update.
Those guidelines were based largely on the series of DECREASE (Dutch Echocardiographic Cardiac Risk Evaluation Applying Stress Echocardiography) studies that reported various cardiac benefits from the use of perioperative β-blockers. DECREASE I found that perioperative bisoprolol reduced short- and long-term cardiac death and MI; a follow-up DECREASE IV found that bisoprolol also significantly reduced 30-day cardiac death and MI in intermediate-risk patients. Then, in 2011, the studies’ principal investigator, Don Poldermans, MD, PhD, was fired from Erasmus Medical Center in Rotterdam, after an investigation found him guilty of scientific misconduct, including failure to obtain patient consent and logging results that did not match patient medical records. A medical center investigation found that the events in DECREASE IV did not match patients’ medical records, and at least one of the DECREASE trials reported almost entirely fictitious data, according to Darrel P. Francis, MD, professor of cardiology at Imperial College’s National Heart and Lung Institute and the senior author of the new study. Dr. Poldermans resigned from the European Society of Cardiology’s (ESC) Committee for Practice Guidelines.
But because the papers have not been retracted from medical journals, the guidelines have remained, leaving clinicians confused about what to do. “I used to feel so reassured recommending perioperative β-blockade,” Dr. Francis said. “It seemed a fantastic solution to a difficult problem.” But now Dr. Francis recommends the therapy mainly for patients with a high risk for MI who must undergo surgery. He suggested that other clinicians review the paper and guidelines to make an informed decision.
Earlier Signs of Trouble
The benefits of perioperative β-blocker therapy have been questioned in previous studies, such as the PeriOperative ISchemic Evaluation (POISE) trial and another meta-analysis by researchers at Brigham and Women’s Hospital, in Boston, both published in the Lancet in 2008 (371:1839-1847; 1962-1976). But it was the POISE study—included in the current study’s analysis—that served as perhaps the strongest initial salvo against perioperative β-blocker therapy. The study compared extended-release metoprolol with placebo in 8,351 patients with atherosclerosis or at risk for the condition. Patients were given the drug between two and four hours before surgery and took it for 30 days after their operations. As expected, fewer patients who received metoprolol had an MI during the study period (4.2% vs. 5.7%; P=0.0017). However, patients given the β-blocker were significantly more likely to have a stroke (41 vs. 19; P=0.0053) or to die (129 vs. 97; P=0.0317) during the 30 days of follow-up.
The bottom line, according to the researchers: For every 1,000 patients given metoprolol for non-cardiac surgery, 15 would avoid an MI but five would have a stroke and eight would die. In addition, although β-blocker therapy would prevent some cases of atrial fibrillation and cardiac revascularization, it would lead to nearly 100 cases of clinically relevant hypotension and bradycardia for every 1,000 people who received the drug.
POISE co-principal investigator Homer Yang, MD, professor and former chair of anesthesiology at the University of Ottawa, in Canada, said the new study findings are consistent with POISE. “Even in our paper, we stated that people need to be careful with patients and not use β-blockers as ‘holy water’ for everyone,” he said. Dr. Yang said he still recommends β-blockers in the perioperative setting, “but not in the low-risk patient population—that’s where the risk–benefit ratio comes in.
“We have always taken a balanced approach,” he added. “We don’t want to [characterize] β-blockers as evil or permanently condemned. Medicine just isn’t like that.”
POISE used “industrial dosages of β-blockers that most of us would never use,” said Kim Eagle, MD, the A. Walter Hewlett Professor of Internal Medicine and director of the Cardiovascular Center at the University of Michigan, in Ann Arbor, but the current study “does call into question how much value versus harm” they offer. Dr. Yang replied that the seemingly high (100 mg) doses of slow-release, once-daily metoprolol metabolize in the body was similar to thrice-daily lower doses (25 mg) of immediate-release metoprolol. The current meta-analysis also explained that, stating “the POISE trial was therefore not high-dose.”
A Balanced Approach
Other physicians also continue to use β-blockers for perioperative purposes but are doing so more judiciously. Dr. Eagle said β-blockers are indicated “for patients with established coronary artery disease and ischemia based on symptoms or non-invasive testing, for whom no contraindications exist, and where the dose is carefully titrated so as to avoid undue bradycardia and or hypotension. They would be especially indicated for patients having major vascular surgery.” If they are used, they should be started at low doses “and never pushed to the expense of low blood pressure,” he said. “We have to be circumspect in how we dose the medicines because they can cause harm.”
W. Scott Beattie, MD, PhD, the R. Fraser Elliott Chair in Cardiac Anesthesia at the University Health Network in Toronto, Ontario, Canada, said the current findings are similar to previous meta-analyses. “Lost in the ‘controversy’ is that β-blockers do exactly what we thought they would—reduce perioperative MI,” he said. “We need to ask: If the risk of heart attack, the leading cause of postoperative mortality, is lowered, why are more patients dying?” The POISE data, Dr. Beattie noted, identified potential safety issues that may be contributing to that excess mortality, including anemia, blood transfusion, emergent surgery and pre-existing cerebrovascular disease. “It is important for us to figure out how to use these drugs safely. Irrespective of starting β-blockers anew, as was done in these studies, 20% to 40% of surgical patients are chronically taking β-blockers. Recent evidence suggests that these patients also have [an] increased number of strokes.”
Dr. Yang and others are trying to tease out which patients might benefit—or be harmed—the most from perioperative β-blocker therapy. One hypothesis is that patients treated during emergency surgery might not benefit as much as those with scheduled surgeries. “Emergency surgery has always been known to be a higher-risk setting,” Dr. Yang said. “In a subgroup analysis [of POISE] … the survival curves of patients getting metoprolol versus controls in emergency surgery do not separate as nicely as the elective cases.”
He also speculated that “emergent cases possibly are already bleeding more or already have more hypotension, which was identified in our POISE study as risk factors for worse outcomes. Additionally, β-blockers reduce the reserve in cardiac output responses, and so those risk factors in emergency surgery may be exacerbated by the use of β-blockers. That’s a hypothesis—we can’t prove it yet.”
Thus, more clinical trials focusing on β-blocker safety are “urgently required,” Dr. Yang added. One early clue has already emerged: A study he was scheduled to present at the Canadian Anesthesiologists’ Society meeting this summer—the meeting was cancelled because of flooding—showed no significant difference between the two groups of patients. He said he is still analyzing those data (abstract 1653709; Figure).
 Number at risk:
Figure. Event-free survival curves for cardiac complications by surgical urgency and treatment assignment. Reprinted with permission from Homer Yang, MD. |
0 件のコメント:
コメントを投稿