In order to bring you the best possible user experience, this site uses Javascript. If you are seeing this message, it is likely that the Javascript option in your browser is disabled. For optimal viewing of this site, please ensure that Javascript is enabled for your browser.
Did you know that your browser is out of date? To get the best experience using our website we recommend that you upgrade to a newer version. Learn more.

We use cookies to optimise the design of this website and make continuous improvement. By continuing your visit, you consent to the use of cookies. Learn more

IRIS: Why does the implantable cardioverter-defibrillator not improve mortality early after myocardial infarction - Insights from the Immediate Risk Stratification Improves Survival (IRIS) trial


Presenter | see Discussant report

Gerhard SteinbeckGerhard Steinbeck (Germany)

List of Authors:
Gerhard Steinbeck, M.D., Karl Wegscheider, Ph. D., Dietrich Andresen, M.D., Karlheinz Seidl, M.D., Johannes Brachmann, M.D., Ellen Hoffmann, M.D., Darius Wojciechowski, M.D., Zdzislawa Kornacewicz-Jach, M.D., Beata Sredniawa, M.D., Géza Lupkovics, M.D., Franz Hofgärtner, M.D., Andrzej Lubinski, M.D., Marten Rosenqvist, M.D., Eik Vettorazzi, M.Sc., Daniel Becker, M.D., and Jochen Senges, M.D., for the IRIS Investigators


In a prospective European multicenter trial, 898 patients at high risk early after myocardial infarction were randomly assigned to treatment with an implantable defibrillator ICD (n=445) or to medical treatment alone (n=453). Although the risk of sudden cardiac death was reduced by ICD therapy, this effect was offset by an increase in the risk of non-sudden cardiac death so that total mortality was unaffected in IRIS: N Engl. J Med 361 (2009) 1427-1436). The mechanism (s) contributing to the increased risk of non-sudden cardiac death in the ICD group remained unresolved. To address this question, an exploratory analysis is performed, with data extraction and validation process still ongoing so that results below have to be considered preliminary.

Instead of assessing the cumulative risk in IRIS, first the daily risk was estimated using a smoothing algorithm together with confidence intervals to give an impression of the randomness of the result. It can be seen that the daily mortality risk of ICD patients is lower for the first two years, but then goes up and never comes back to the level of the control group. The sudden cardiac death - specific hazard curve reveals that the ICD is quite effective to do what it is expected to do for the first two years, but not thereafter. Finally, the risk of non-sudden cardiac death in the ICD group is higher than in the control group during the whole follow-up period, this difference being most readily apparent at three years of follow-up.

In multivariate analyses with 30 baseline characteristics, 7 covariates were found to be independent predictors of death in the total IRIS population, however only one cause–specific ICD subgroup effect was noted: the ICD did not reduce sudden cardiac death and did not increase non-sudden cardiac death in a subgroup of 91 STEMI patients without reperfusion, as opposed to IRIS patients with reperfusion and to the total IRIS population (p<0,001). As a result of this divergence, the 91 STEMI patients without reperfusion were eliminated from further exploratory analysis.

Finally, the baseline model was extended by including time-dependent covariates Episodes and ICD interventions, taken from the readings of the device records of the ICD patients and adjudicated by an independent board of experts, are modelled as intermediate events indicating a potential risk. According to this model any appropriate therapy of the ICD (n=88 of 375 = 23,5%), compared to control, is associated with increased total mortality due to elevated non-sudden cardiac death rate, while there is no statistically significant effect associated with appropriate shocks only, inappropriate shocks only, and antitachycardia pacing only. Periods with both appropriate and inappropriate shocks are at especially high risk of death (hazard ratio 4,7) and non-sudden cardiac death (hazard ratio 9,9) probably reflecting a shift from sudden to non-sudden cardiac death.

Lastly, an increase of % right ventricular antibradycardia pacing by 10 (assessed from last interrogation versus overall mean) was associated with a significant increase of total mortality and non-sudden cardiac death, an effect independent of anti-tachycardia ICD interventions.

Discussant | see Presenter abstract

Christophe Leclercq (France)

Presentation webcast

Presentation slides


The current European and North-American guidelines for ICD implantation for primary prevention exclude patients with recent myocardial infarction (MI) i.e within the 40 first days after the index MI. However, we know from different studies that the risk of SCD is significantly higher just after a MI especially in patients with a low LV ejection fraction.

Two trials, the DINAMIT and the IRIS trial were conducted to assess the efficacy of an early implantation of an ICD after a MI and their results published in 2004 and 2009 were concordant showing a reduction in sudden cardiac death (SCD) and an increase in non SCD, resulting in a neutral effect on total mortality. Dr Steinbeck and colleagues present a new analysis of the IRIS data different statistical tools.

The results did show that ICD reduced SCD only during the first two years after implantation and was associated with a constant increase in non SCD especially after 3 years of follow-up. As observed previously in the DINAMIT trial age, multivessel coronary artery disease, NHYA class III or IV and non optimal drug treatment were independent predictors of mortality.
Interestingly ICD did not reduce SCD in the small sub-group of patients (n= 91) without reperfusion. The presence of RV pacing was associated with an increase in non SCD and total mortality. Finally iCD intervention, appropriate or not, were associated with an increase in non SCD.

An interesting analysis form the DINAMIT trial showed that patients with ICD and appropriate therapies had more intercurrent events especially myocardial ischemia and heart failure during the follow-up. Ventricular arrhythmias even if they are treated by ICD intervention may enhance cardiac deterioration but in some cases an intercurrent event, particularly an ischemic event, may generate ventricular arrhythmias which are an epiphenomenon of the cardiac deterioration. Finally, preliminary data have shown that shocks may directly or indirectly increase mortality suggesting the extension of the ATP programming even in the VF zone.

These preliminary data from this new analysis did not answer the question of why does the implantable cardioverter-defibrillator not improve mortality early after a myocardial infarction. So today we have no evidence to modify the current guidelines


710005 - 710006


Clinical Trial Update II - Rate and rhythm

The content of this article reflects the personal opinion of the author/s and is not necessarily the official position of the European Society of Cardiology.