NONINVASIVE ELECTROPHYSIOLOGY
Risk stratification after myocardial infarction: a new method of determining the neural component of the baroreflex is potentially more discriminative in distinguishing patients at high and low risk for arrhythmias
1 Department of Cardiology, Division of Clinical Electrophysiology, J.W. Goethe University, Frankfurt, Germany; 2 Medical University Lodz, Lodz, Poland; 3 Cardiology Practice Bonn, Im Mühlenbach 2B, Bonn D-53127, Germany; 4 Medtronic Bakken Research Center, Maastricht, The Netherlands
Aims: We hypothesize that the neural component (NC) of the baroreflex sensitivity (BRS) is a better risk stratifier for ventricular tachycardia/fibrillation (VT/VF) than conventional BRS itself, because it is both independent of vessel wall stiffness and can be measured non-invasively.
Methods and results: NC was determined by correlating spontaneous carotid artery diameter variations with R–R interval variations using spectral analyses. In consecutive outpatient populations with chronic coronary artery disease the ability of the NC to distinguish post-myocardial infarction (MI) patients at risk for VT/VF (post-MIHIGH RISK) from post-MI less prone to arrhythmias (post-MILOW RISK) was compared with the pressure-derived BRSphenyl and BRSspectral method. Ninety-six patients, i.e. 28 post-MILOW RISK, 28 post-MIHIGH RISK [a LVEF(left ventricular ejection fraction) <30% and/or history of VT/VF] and 40 healthy controls were enrolled. With NC, rather than with BRS methods, median values for post-MIHIGH RISK were smaller than for post-MILOW RISK patients (NC, P = 0.03; BRSspectral, P = 0.35; BRSphenyl, P = 0.63). Variability of R–R interval (LF = 0.04–0.15 Hz) was significantly larger in the control group than in the post-MIHIGH RISK and post-MILOW RISK group (P < 0.01 and P < 0.01). To separate post-MIHIGH RISK from post-MILOW RISK patients, a linear combination of age and the logarithm of the NC measurement was constructed as a risk index. By optimizing the intercept of this line, an optimal sensitivity and specificity pair was determined. The sum of optimal specificity and sensitivity was higher for NC (155) than for BRSspectral (133) and BRSphenyl method (132). With all methods, values for post-MI patients were significantly smaller than for controls.
Conclusion: NC may be superior to conventional BRS measures in identifying post-MI patients at high risk for VT/VF.
Key Words: Baroreflex, Neural component, Risk stratification
* Corresponding author. Tel: +49 228 623324; fax: +49 228 616881.E-mail address: klingenheben{at}bonn-kardiologie.de; klingenheben{at}aol.com
The first 2 authors contributed equally to the study.
Manuscript submitted 25 September 2007. Accepted after revision 20 December 2007.