CARISMA: Markers of autonomic function stratify post-MI, low-LVEF sudden-death risk

May 16, 2007

Denver, CO - Noninvasive markers of autonomic dysfunction, especially when combined with ECG criteria, are at least as predictive of sudden-death risk as positive electrophysiologic (EP) studies in patients with a low LVEF after an MI, suggests a prospective study reported here last week at the Heart Rhythm Society (HRS) 2007 Scientific Sessions[1]. According to its investigators, the noninvasive markers could potentially find a role in identifying post-MI patients who are most likely to benefit from an implantable cardioverter defibrillator (ICD).

Also in the trial, called Cardiac Arrhythmias and Risk Stratification after Myocardial Infarction (CARISMA), insertable loop recorders (ILR) worn by the 297 patients documented frequent, largely asymptomatic spontaneous arrhythmias or other conduction disturbances, including atrial fibrillation, in almost a third of the population over a mean follow-up of about two years. A tenth of the group developed ventricular fibrillation (VF) or sustained ventricular tachycardia (VT) and about one eighth had episodes of nonsustained VT, observed Dr Heikki V Huikuri (University of Oulu, Finland) when presenting the results of CARISMA, which was carried out at 10 centers in Belgium, Denmark, Finland, the Netherlands, and Norway.

Several noninvasive markers as well as VT/VF inducibility were significantly predictive of the study's composite primary end point, which was sudden arrhythmic death, resuscitated arrhythmic death, or spontaneous sustained VT—conditions considered likely to be treatable with an ICD, according to Huikuri. The strongest predictor of the primary end point was a combination of two noninvasive markers, one a reflection of abnormal arrhythmia substrate and the other a measure of autonomic dysfunction, he observed. A ratio of the QRS duration by signal-averaged electrocardiography (SAECG) to an index of heart-rate variability (HRV) that reached or passed a certain threshold indicated an eightfold increase in risk of the primary end point.

"If this ratio was abnormal, one out of three patients experienced a VT or VF event," Huikuri said. "This actually may [potentially] become a helpful tool for identifying patients who will benefit from an ICD."

I'm not sure we can convince the medical community that everyone should be tested for heart-rate variability at six weeks post-MI.

CARISMA's patients were provided with the ILR devices five to 21 days after experiencing an acute MI; they were required to have an LVEF <40, and those in NYHA class 4 heart failure or for whom CABG was planned were excluded, Dr Paul Erik Bloch Thomsen (University of Copenhagen, Denmark) said when presenting the study's methods. At discharge, 71% were taking the triad of beta blockers, statins, and either an ACE inhibitor or angiotensin receptor blocker.

Six weeks postinfarction and every three months for two years, the patients underwent EP programmed stimulation studies and exercise testing, 24-hour Holter monitoring (to disclose HRV and any preventricular contractions or nonsustained VT), SAECG (to measure QRS duration), 12-lead ECG (for QT dispersion and QRS duration), and T-wave alternans (TWA) testing.

Rhythm disturbances were common, observed Huikuri. In all, 62% experienced one or more arrhythmias, although nine times out of 10 they were asymptomatic. About 7% of the patients had bradycardia, 10% had high-grade atrioventricular block, 13% had nonsustained VT, 5% experienced sinus arrest, 32% had atrial fibrillation, 3% had sustained VT, and 3% had VF.

The primary end point was reached by 8.7% of the patients, Huikuri reported; the rate of sudden cardiac death alone was 3.8%.

Of the purported risk markers measured at six weeks, LVEF, nonsustained VT, both 12-lead ECG criteria, and—"perhaps surprisingly," said Huikuri—TWA failed to significantly predict the primary end point. Emerging as significant predictors, however, were the two HRV indexes (including the standard deviation of all normal RR intervals [SDNN]), a QRS >120 ms on the SAECG, and the ratio of the SAECG-QRS to SDNN.

CARISMA: Prognostic power of autonomic and electrophysiology measures of sudden-death risk

Parameter HR (95% CI) PPV (%) NPV (%) Sensitivity (%)
Heart rate variability indexes        
SDNN 4.4 (1.6-11.8) 21 95 35
VLF frequency spectral component 6.4 (2.4-16.8) 26 95 41
SAECG QRS duration >120 ms 4.5 (1.8-11.4) 20 95 85
VT/VF inducibility at EP testing 3.9 (1.5-10.2) 14 96 53
SAECG QRS duration /SDNN >1.6 8.3 (3.1-22.3) 33 95 44
PPV=positive predictive value
NPV=negative predictive value
SDNN=standard deviation of all normal RR intervals
VLF=very low frequency
SAECG=signal averaged electrocardiography

During the CARISMA presentation's question-and-answer period, session comoderator and HRS president Dr Dwight W Reynolds (University of Oklahoma Health Sciences Center, Oklahoma City) cautioned that the findings are based solely on risk stratification six weeks after MI and asked whether data obtained at any other time were also prognostic. In response, Huikuri said that Holter monitoring at one week post-MI was the only test performed at other than six weeks, and its HRV readings were not as predictive.

Asked by the other comoderator, Dr Hein J Wellens (University Hospital Maastricht, the Netherlands), how the results may have influenced his own clinical practice, Huikuri said, "I'm not sure we can convince the medical community that everyone should be tested for heart-rate variability at six weeks post-MI." However, he added, if asked which tests he'd do under the classic hypothetical circumstance of the patient being his wife, he said he would do "all of them," including TWA, and then decide whether an ICD was called for.

CARISMA was sponsored by Cambridge Heart and the Medtronic Bakken Research Center. Potential conflict-of-interest relationships were reported by Thomsen for Medtronic, Boston Scientific, and St Jude, and by Huikuri for Medtronic and Biotronik. Wellens reports receiving consulting fees or honoraria from the Medtronic Bakken Research Center. Reynolds reports receiving consulting fees or honoraria, serving on the speakers' bureau for, and having equity interest in or stock options with Medtronic.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.