ACC: MRIs can determine when to implant high-risk patients with ICDs
Infarct size measured by MRI can be used as a predictor of ventricular tachycardia and may answer whether or not high-risk patients with an ejection fraction under 35 percent should be implanted with implantable cardioverter-defibrillators (ICDs), said Jeffrey J. Goldberger, MD, from Northwest Memorial Hospital in Chicago, during a presentation at ACC.10 in Atlanta March 15.
When beginning to answer whether or not ICDs should be placed in patients with an ejection fraction under 35 percent, Goldberger said, “What we have to look at is what is the evidence, what are the benefits and what are the risks.”
Referencing the MADIT II trial by Moss et al, he said the results showed no reduction in the risk of sudden cardiac death in patients implanted with ICDs and at high risk.
During Goldberger’s research, he and colleagues assessed dialysis patients who had already been implanted with ICDs, versus non-dialysis patients. Results showed that after almost four years, half of the dialysis patients were “already dead.” After these poor results, he said that you must ask: “How much benefit can there be for an ICD if half the patients are dead, even with an ICD?”
Goldberger said that this question is often difficult to answer due to the fact that no previous studies have evaluated dialysis patients or patients who have an ejection fraction below 35 percent. “I will argue to you that the benefit of ICDs in dialysis patients is at best marginal. It's unknown, but it is at best marginal,” he said. “We have no data yet to know if there is any benefit for the ICD in this patient population.”
While there are still no statistics to determine the potential benefit of the use of ICD for these patients, when looking at the competing risks, dialysis patients had a high competing risk for death from other causes, Goldberger said.
“This could be possibly why patients have less of a benefit from an ICD, whereas your standard patient with an ejection fraction above 35 percent has a low competing risk and has a great potential to benefit for the ICD,” he explained.
“If we implant a lot of ICDs in dialysis patients, we are going to have very low impact on reducing the number of sudden cardiac deaths,” Goldberger said. “If we find the right population of patients with ejection fractions above 35 percent, we have a potentially very high impact on reducing sudden cardiac death.”
Goldberger said that two questions must be taken into consideration: How big is the population of those who exhibit an ejection fraction above 35 percent and at risk for sudden cardiac death, and how do we identify these patients at risk?
Goldberger referenced a report conducted by the multi-center post-infarction diltiazem research group who found a strong relationship between ejection fraction and mortality. The research group found that as ejection fractions decrease, mortality rises. “In essence I think this is consistent with our clinical experience,” he said. “Lower ejection fractions result in higher mortality rates."
“The number of events that occur in high-risk patients is much smaller and we have to start to target these populations if we want to have any impact on the epidemiology of sudden cardiac death,” said Goldberger.
Researchers from Northwest Memorial Hospital looked at the measurements of infarct size on patients who were inducible, compared to those who were not. Those “patients who were not inducible had much smaller infarct sizes,” he said.
Currently, delayed-enhanced MRI can help to evaluate infarct size, Goldberger noted. “This is very important because ventricular arrhythmias are related to the substrate which is the MI and ejection fraction.” This can help analyze the infarction itself rather than the substrate, he explained.
“Ejection fraction is not really a good surrogate for the substrate for ventricular arrhythmias,” said Goldberger. He pinpointed ejection fraction as being “only a very poor man’s measure of infarct size and that’s why it’s probably not a very good predictor of ventricular arrhythmias.”
In a study of 857 patients who underwent MRI, researchers were able to quantify delayed enhancements for two groups, one with ejection fractions of above 50 percent and the rest below 50 percent. According to Goldberger, “Just the presence of infarct size affects your survival. This turns out to be a very powerful tool to assess for ventricular arrhythmia risk.”
While Goldberger said that MRI is “not the only tool that can be used for assessing risk of sudden cardiac death in patients with an ejection fraction above 35 percent,” one must evaluate infarct size in order to help predict arrhythmia risk. “It is probably the best test to identify substrate,” he said. However, he noted that future randomized clinical trials will be necessary to prove this assessment.
During the presentation, Douglas P. Zipes, MD, from Krannert Institute of Cardiology in Indianapolis argued that infarct size was "simply another substitute of a surrogate for the electrophysiologic event causing sudden death.” He explained that what "this does is look at the anatomy and say that ventricular tachycardia/ventricular fibrillation may be better than an ejection fraction, but still does not look at the primary event.”
When beginning to answer whether or not ICDs should be placed in patients with an ejection fraction under 35 percent, Goldberger said, “What we have to look at is what is the evidence, what are the benefits and what are the risks.”
Referencing the MADIT II trial by Moss et al, he said the results showed no reduction in the risk of sudden cardiac death in patients implanted with ICDs and at high risk.
During Goldberger’s research, he and colleagues assessed dialysis patients who had already been implanted with ICDs, versus non-dialysis patients. Results showed that after almost four years, half of the dialysis patients were “already dead.” After these poor results, he said that you must ask: “How much benefit can there be for an ICD if half the patients are dead, even with an ICD?”
Goldberger said that this question is often difficult to answer due to the fact that no previous studies have evaluated dialysis patients or patients who have an ejection fraction below 35 percent. “I will argue to you that the benefit of ICDs in dialysis patients is at best marginal. It's unknown, but it is at best marginal,” he said. “We have no data yet to know if there is any benefit for the ICD in this patient population.”
While there are still no statistics to determine the potential benefit of the use of ICD for these patients, when looking at the competing risks, dialysis patients had a high competing risk for death from other causes, Goldberger said.
“This could be possibly why patients have less of a benefit from an ICD, whereas your standard patient with an ejection fraction above 35 percent has a low competing risk and has a great potential to benefit for the ICD,” he explained.
“If we implant a lot of ICDs in dialysis patients, we are going to have very low impact on reducing the number of sudden cardiac deaths,” Goldberger said. “If we find the right population of patients with ejection fractions above 35 percent, we have a potentially very high impact on reducing sudden cardiac death.”
Goldberger said that two questions must be taken into consideration: How big is the population of those who exhibit an ejection fraction above 35 percent and at risk for sudden cardiac death, and how do we identify these patients at risk?
Goldberger referenced a report conducted by the multi-center post-infarction diltiazem research group who found a strong relationship between ejection fraction and mortality. The research group found that as ejection fractions decrease, mortality rises. “In essence I think this is consistent with our clinical experience,” he said. “Lower ejection fractions result in higher mortality rates."
“The number of events that occur in high-risk patients is much smaller and we have to start to target these populations if we want to have any impact on the epidemiology of sudden cardiac death,” said Goldberger.
Researchers from Northwest Memorial Hospital looked at the measurements of infarct size on patients who were inducible, compared to those who were not. Those “patients who were not inducible had much smaller infarct sizes,” he said.
Currently, delayed-enhanced MRI can help to evaluate infarct size, Goldberger noted. “This is very important because ventricular arrhythmias are related to the substrate which is the MI and ejection fraction.” This can help analyze the infarction itself rather than the substrate, he explained.
“Ejection fraction is not really a good surrogate for the substrate for ventricular arrhythmias,” said Goldberger. He pinpointed ejection fraction as being “only a very poor man’s measure of infarct size and that’s why it’s probably not a very good predictor of ventricular arrhythmias.”
In a study of 857 patients who underwent MRI, researchers were able to quantify delayed enhancements for two groups, one with ejection fractions of above 50 percent and the rest below 50 percent. According to Goldberger, “Just the presence of infarct size affects your survival. This turns out to be a very powerful tool to assess for ventricular arrhythmia risk.”
While Goldberger said that MRI is “not the only tool that can be used for assessing risk of sudden cardiac death in patients with an ejection fraction above 35 percent,” one must evaluate infarct size in order to help predict arrhythmia risk. “It is probably the best test to identify substrate,” he said. However, he noted that future randomized clinical trials will be necessary to prove this assessment.
During the presentation, Douglas P. Zipes, MD, from Krannert Institute of Cardiology in Indianapolis argued that infarct size was "simply another substitute of a surrogate for the electrophysiologic event causing sudden death.” He explained that what "this does is look at the anatomy and say that ventricular tachycardia/ventricular fibrillation may be better than an ejection fraction, but still does not look at the primary event.”