Background: The relationship between ventricular arrhythmias (VA) in the subacute phase of a myocardial infarction (MI) and subsequent major arrhythmic events, i.e. sustained ventricular tachycardia (SVT) and sudden death (SD), is well known. The importance of left ventricular dysfunction in the same context is also well established. The vast majority of the data in the literature come from the prefibrinolytic era and/or are derived from limited data bases.
Materials, methods and results: We examined the large, uniform GISSI-2 population of acute fibrinolysed myocardial infarctions in order to evaluate the significance and predictive power of VA detected by Holter monitoring during the subacute phase. Particular attention was paid to the occurrence of nonsustained ventricular tachycardias (NSVT) since their role is still uncertain, so it is hard to assess the utility of invasive measures such as programmed electrical stimulation (PES). Left ventricular function was evaluated by mono-, and two-dimensional echocardiography. Holter monitoring was available in 8,676 patients; a six-month follow-up, as regards mortality was completed in 8,552 patients (98.5%) and, as regards SVT incidence, in 7.713 (88.9%). During the follow-up 256 patients died (3%), 84 out of them suddenly (1%). Twenty-six [corrected] patients out of 7,713 (0.3%) experienced one or more SVT. The relationship between VA, left ventricular dysfunction and major arrhythmic events was evaluated by odds ratios (OR) and their confidence intervals (CI). Odds ratios for the combined end-point (SD and/or SVT) was 4.49 (CI 95% 2.69-7.51) if the Holter showed > 10 VEB/hour; 2.34 (CI 1.48-3.68) if couplets were present; 3.26 (1.52-6.99) if NSVT were present; 3.02 (2.02-4.50) if any complex ventricular arrhythmia was detected. Left ventricular disfunction seemed to exert a more powerful influence: OR 9.80 (CI 5.75-16.69) for SD and/or SVT. Any arrhythmic parameter had very low positive (< or = 3%) and very high (approximately 99%) negative predictive power. Multivariate analysis (Cox Model) including major prognostic factors confirmed the independent prognostic value of frequent VA (RR 2.75; 95% CI 1.58-4.79), couplets (RR 1.91; CI 1.28-2.86); complex VA (RR 2.02; CI 1.36-3.00). NSVT, after adjusting for the selected risk factors, was not independently associated with a worse prognosis.
Conclusions: Ventricular arrhythmias detected by Holter monitoring during the subacute phase of a MI still have prognostic significance for major arrhythmic events in the fibrinolytic era. The presence of NSVT triples the risk of SD and/or SVT in the six months after an acute MI, but loses any predictive power in a multivariate analysis. Only 6.6% of the patients showed one or more episodes of NSVT in the Holter recording. If the ongoing clinical trials, MUSTT and MADIT, will confirm the usefulness of PES in such patients, the benefit will be confirmed to a very small proportion of the patients at risk of specific diagnostic tests. The positive predictive power of VA is very low and it is therefore mandatory to add other non-invasive variables to the screening to identify the subgroups at greatest risk. On the other hand the very high (99%) negative predictive power of VA and left ventricular dysfunction enables us to identify a large population of infarcts with a negligible risk, who need no further sophisticated investigations. From the point of view of the cost/benefit evaluation this seems to be an outstanding result.