Medicine is not an exact science, however, a significant amount of effort is put everyday in research endeavors directed toward minimizing human or system's errors and improve safety. Large randomized trials appear every day in the different journals around the globe and in fact, can change the whole practice of Medicine in one way or the other; a critical review of the data is recommended, but given the impressive amount of literature that appears everyday, people may just go ahead with the conclusions of the studies without further examining the "fine print in the text". Every one percent in mortality or survival is important as it affects a large population.Dogmatic teaching has been replaced by a mixture of experience, evidence and common sense; we become pragmatic in the way we use knowledge as we need to be extremely cautious when examining new data before considering it as the "real truth". Controversy arises when a previously established dogma (not necessarily a mistake of the past) appears to be severly questionned by new evidence, putting in perspective the frailty of our belief system. What is truth today, may be a lie tomorrow. Medicine evolves, so does the methodology in research as well as the amount of knowledge that supports new research endeavors. The main issue is that meanwhile this turmoil of data and knowledge, lie or truth, happens, patients are subjected to medical care based on the current evidence.
Since the last decade, data came up supporting the use of betablockers in the perioperative setting since Dr. Mangano's trial with atenolol, (NEJM 1996). Posteriorly Poldermans (NEJM, 1999) published the initial data using bisoprolol in high risk surgery patients in the DECREASE study. Multiple back and forth discussions occurred after a dramatic increase in perioperative betablocker use and in 2005, Lindenauer (NEJM) demonstrated that only high risk patients (essentially those with RCRI > 2) had a survival benefit from the perioperative use of betablockers. Most recently, the POISE trial (Devereaux, Lancet, 2008) showed an impressive amount of adverse complications in patients using perioperative betablockers, especially increased risk for stroke and sepsis; however its methodology was controversial as at differing from DECREASE in which low dose of betablocker was used initially with careful uptitration over a period of 30 days, at POISE a large dose (200 mg) of metoprolol succinate was used immediately prior to surgery.
Yesterday, Dr. Poldermans and its group published the most recent data on the DECREASE IV trial, and as in their previous publications, showed an impressive effect on survival - this time with an almost 70% decrease in 30 days mortality in patients using bisoprolol.
I have several comments on the study. I am amused by his positive results which he (Dr. Poldermans) has always strongly defended. Here are my thoughts:
1. As already known, DECREASE is an open labeled study (as that was the only way they could titrate up the betablocker), so although it was randomized, it was not blinded and this can yield to treatment bias which they recognize as on of their limitations.
2. They only recruited 1066 patients (from an original goal of 6000 (1500 per group) to detect an the anticipated risk reduction of 30% with a power of 81% and a 2-sided alpha of 5%). - however given the small number of patients in comparison (16% of the original goal) they obtained a dramatic result in 30 days mortality, however although the proportionally smaller sample did not affect negatively the betablocker group, it presumably had an effect on the statins group which had a trend toward decrease in mortality, although not statistically significant.
3. No seconary end-points were included in this article such as stroke, sepsis, etc....which makes POISE stronger methodologically speaking. However the authors make the point that the POISE trial showed a 1% incidence of stroke in the group randomized to metoprolol compared with 0.5% in the control group. In comparison, the incidence of stroke was 0.4% in the DECREASE studies, with no difference between groups.
4. The discussion provided is solid making the arguments against the unfavorable results from POISE. States that the beneficial effect of beta-blockade on coronary plaque stability, related to sustained mechanical and antiinflammatory effects require weeks to develop.
5. DECREASE IV has focused on the INTERMEDIATE risk group of patients, which do not take into account the low risk, however, the Lindenauer paper from NEJM in 2005 showed deleterious effect of betablockers in persons with RCRI score <2.> 3 carefully starting a low dose at the beginning (atenolol or metoprolol succinate 12.5 mg per day) with slow up-titration to a HR of 60 to 70.
In summary, we need to be very careful. Perhaps only patients with RCRI 3 or more really benefit from it and a very low dose should be started. Ideally slow and careful uptitration should be achieved. Realistically, we need to optimize patients almost immediately prior to surgery, therefore, starting a low dose is advised and titrate it up post-operatively.
1. Dunkelgrun M, et al. Bisoprolol and Fluvastatin for the Reduction of Perioperative
Cardiac Mortality and Myocardial Infarction in Intermediate-Risk Patients Undergoing Noncardiovascular Surgery A Randomized Controlled Trial (DECREASE-IV). Ann Surg. 2009;249: 921–926.
2. Auron M. Perioperative Betablockers and the POISE: Evidence revisited. Slideshare, http://bit.ly/hceO6