We congratulate Dr. Mahla et al.  1on their recent work on using brain natriuretic peptide (BNP) to predict adverse cardiac outcomes after vascular surgery. We would like to comment on the editorial by Augoustides and Fleischer2on the possible future use of BNP in this regard.

It is obviously highly desirable to now move from using BNP to better predict risk to using it to better reduce risk. However, this next step requires careful thought because, unlike many other risk predictors, BNP is not a culprit itself but is rather a marker of other culprits. We feel that the next crucial step is to fully phenotype high-BNP/high-risk preoperative individuals to better target the “multimodal perioperative interventions” suggested by Augoustides and Fleischer.

This phenotyping step is crucial because BNP can be elevated by virtually any cardiac abnormality, many of which will be silent. BNP can be elevated by (silent) myocardial ischemia, valve disease, left ventricular dysfunction, left atrial dilatation, left ventricular hypertrophy, or even atrial fibrillation. Many of these may coexist in some individuals. The study of Mahla et al.  1only excluded a few of these causes of elevated BNP (left ventricular systolic dysfunction, atrial fibrillation, and one form of valve disease). Therefore, many of the patients studied who had high BNP could harbor silent ischemia, left atrial dilatation, left ventricular hypertrophy, and subtle valve disease other than aortic stenosis. Surely a crucial step is for us to fully phenotype such high-BNP/high-risk individuals to get a full picture of the spectrum of their underlying cardiac abnormalities so that when we devise intervention trials to try to reduce their risk, the interventions will be targeted to the underlying cardiac abnormality, i.e. , “smarter” trials. This is because each of the different aforementioned cardiac abnormalities (if found) would best be treated in different ways, e.g. , β blockade, higher statin dose or angioplasty for silent ischemia, angiotensin receptor blocker, aldosterone blockade and ultralow blood pressure for left ventricular hypertrophy, ARB and anticoagulation for left atrial dilatation, and even valve repair or valvuloplasty in some valve disease cases. Surely we ought not to devise therapeutic interventions before we fully phenotype these individuals to better understand the source of their increased BNP/increased risk.

Phenotyping high-BNP preoperative individuals could produce another benefit. This is because experience in other settings suggests that an obvious cause for the high BNP may not be found in every case. Therefore, phenotyping high-BNP individuals may enable us to even better target our interventions because we could then target our interventions to those who fail two high-risk screening tests rather than just one, i.e. , therapies could be targeted not just to those who failed a BNP test but to those who failed a BNP test and a search for silent cardiac abnormalities.

That is, high BNP patients may need to be phenotyped not only because the information will be necessary for us to move forward to “smarter” trials, but also because it may be more cost effective ultimately to target whatever therapies are devised toward those who fail two screening tests (BNP and phenotyping) rather than just one (BNP).

*Ninewells Hospital and Medical School, Dundee, United Kingdom. c.c.lang@dundee.ac.uk

Mahla E, Baumann A, Rehak P, Watzinger N, Vicenzi MN, Maier R, Tiesenhausen K, Metzler H, Toller W: N-terminal pro-brain natriuretic peptide identified patients at high risk for adverse cardiac outcome after vascular surgery. Anesthesiology 2007; 106:1088–95
Augoustides J, Fleisher LA: Advancing perioperative prediction of cardiac risk after vascular surgery: Does postoperative N-terminal pro-brain natriuretic peptide do the trick? Anesthesiology 2007; 106:1080–2