14th International Congress
THE "NEW FRONTIERS"
OF ARRHYTHMIAS 2000

Jan. 29 - Feb. 5, 2000
Marilleva, Trento, Italy

RT-144

An update on clinical trials for the primary prevention of sudden cardiac death

Adam Zivin, Gust H. Bardy.
University of Washington Medical Center, Seattle, USA

Appropriate targets for primary prevention

The first question to ask is who will benefit most from prophylactic treatment of ventricular tachyarrhythmias. Coronary disease is found in approximately 80% of sudden death survivors, but cardiac arrest is often the first manifestation of the disease in a person thought to have been healthy^9,10. At present, there is no means of screening the general population for risk of sudden death. Furthermore, the lower the risk in the population under study, the larger the trial will have to be to demonstrate a mortality benefit. Therefore, to address the relative value of therapies for primary prevention, the focus should be on patient groups presumed to have a higher incidence of ventricular arrhythmias; those with known structural heart disease.

The largest such group at risk is patients with dilated cardiomyopathy, whether ischemic or non-ischemic in etiology. Mortality at 2 years ranges from 12-15% in asymptomatic patients to 70-75% in Class IV heart failure^11,12. Death among patients with left ventricular dysfunction is both sudden and non-sudden. Despite a better prognosis, those who are less functionally impaired are more likely to have a death classified as “sudden”¹³. A majority (90-95%) of all patients with cardiomyopathy will have some form of ventricular arrhythmia ranging from PVCs to sustained VT, or VF^14-19. The incidence of non-sustained VT (NSVT) increases with worsening functional class; from 15-20% in Class I-II to 50-70% in Class III-IV and the majority of these arrhythmias are asymptomatic^20,21.

Post myocardial infarction patients are also known to have an increased risk of ventricular arrhythmias. The incidence of NSVT in the peri-infarct period is typically 5-7%, with sustained VT or VF observed in 0.3-3%, and continued complex ectopy reported in roughly 20% of patients at three weeks^22-24. While it is tempting to equate these non-sustained arrhythmias with “non-sustained sudden death”, the available data does not support this intuitive link. Not only does the presence of NSVT or PVCs fail to adequately predict sudden death; its value in predicting total mortality has been questioned²⁵. The MILIS investigators concluded that the presence of >10 PVCs/hr was an independent marker of poor prognosis, however 49% of the patients in the group with highest mortality (³10 PVCs/hr and EF£40%) were on treatment with class I antiarrhythmic agents compared with only 12% of those with the lowest mortality (<10 PVCs/hr and EF>40%). In a pre-thrombolytic study of post-infarct patients, the highest mortality was observed in those with an ejection fraction £40% and “complex” ventricular ectopy (Lown class III-V)²⁶. Again, 62% of those who died during follow-up were on treatment with a class I agent compared with 40% of the survivors. Of note, reflecting the presumptions of the time, the authors of this study speculated that the fact that only 18 of 29 patients with complex ectopy were treated with antiarrhythmic drugs was a reflection of inadequate methods for detection, and thus inadequate treatment, of life-threatening arrhythmias during hospitalization. In a study of patients treated with amiodarone for VT or VF post-MI, however, a similar conclusion regarding the prognostic significance of NSVT was reached²⁷. Recent studies have come to divergent conclusions. Analysis of CHF-STAT found that NSVT did not predict mortality independent of functional class and EF²⁰. The GESICA study, which randomized primarily non-ischemic patients, came to the opposite conclusion²⁸. Given the available data, the presence of PVCs or NSVT on Holter monitoring in patients with cardiomyopathy does not appear to provide any additional value in prediction of future sustained arrhythmias.

The combined incidence of sustained ventricular arrhythmias and sudden, presumably arrhythmic death in infarct survivors and patients with known coronary disease ranges from 11 to 66% at 2 years with highest total mortality observed in those with concomitant congestive heart failure^12,15,18. Given that the risk of sudden death is highest in the first few months after infarction, primary prevention would be most effective if the highest risk patients could be identified during the index hospitalization²⁹. Paradoxically, in contrast to the significance assigned to PVCs and NSVT, conventional wisdom has held that sustained VT or VF in the first 72 hours following an acute MI contributes to in-hospital mortality, but does not prognosticate future events^19,30-32. Two recent studies have challenged this assumption. Based on retrospective analysis of the GUSTO database, while early (<48°) VF was not a useful marker, the occurrence of VT at any point, or VT and VF during the index hospitalization conferred a worse 1-year mortality²³. Similarly, a retrospective single-center study of MI patients concluded that NSVT occurring later than 13 hours after presentation marked a higher 1 year total mortality²². As with cardiomyopathy, in post-MI survivors support for non-sustained ventricular arrhythmias as an independent predictor of presumably arrhythmic death has not been convincingly established. Whether sustained monomorphic VT will prove helpful in targeting post-MI patients for prevention of arrhythmic death requires further study.

Beyond Holter monitoring, further risk stratification for patients with coronary disease, and ischemic and non-ischemic cardiomyopathy has been attempted with numerous techniques, including: invasive electrophysiologic study (EPS), signal averaged ECG, heart rate variability, T-wave alternans and QT dispersion among others^33-43. SAECG has been reported as having a negative predictive value of 90-95% for various combined endpoints, but as with the other methods, the positive predictive value has been low. While often used, EPS has lacked any placebo-controlled data supporting its validity in risk stratification. Ongoing post-hoc analysis of the MUSTT database may help clarify this issue.