Introduction The occurrence of microvolt T-wave alternans (MTWA) at low heart rates has been suggested as a useful predictor of ventricular arrhythmia (VA) in patients with heart failure. However, results of recent clinical studies have questioned the specificity of MTWA for VA, and call for a better understanding of the link between alternans and VA. Cellular repolarisation alternans underlies MTWA and discordant repolarisation alternans has been experimentally linked to the production of VA. We hypothesised that increased heterogeneity of action potential duration in heart failure would lead to a greater vulnerability to discordant alternans and so to VA.
Methods Adult male New Zealand white rabbits (n = 11) underwent surgical coronary arterial ligation to produce myocardial infarction. Hearts from sham-operated (n = 4) and normal (n = 13) rabbits were used as controls. Eight weeks after surgery rabbits were killed with an intravenous injection of sodium pentobarbitone (100 mg/kg). Hearts were excised and arterially perfused with oxygenated Tyrode’s solution maintained at 30°C (n = 16) or 37°C (n = 15). Left ventricular wedge preparations were loaded with RH237 and an optical mapping system was used to record transmural optical action potentials during a progressive shortening of pacing cycle length. Custom written analysis software was used, and data are expressed as mean ± SEM.
Results Repolarisation alternans, confirmed by spectral analysis, was observed during rapid pacing in all hearts at 30°C. Alternans magnitude increased with increasing heart rate (HR). The HR threshold for alternans was lower in failing hearts (294 ± 14 vs 337 ± 7 bpm, p<0.01). At 37°C, repolarisation alternans was significantly more common in failing hearts (5/7 vs 1/8, p<0.05) and the HR threshold for alternans was also significantly lower (396 ± 18 vs 449 ± 14 bpm, p<0.01). Transmural discordant alternans was observed in three of 17 normal hearts and one of 14 failing hearts (p = NS) and was not seen at 37°C. The induction of VA during rapid pacing was significantly more common in failing hearts at each temperature (2/9 vs 6/7 at 30°C, 1/8 vs 5/7 at 37°C, p<0.05 for each). Transmural alternans was present directly before the induction of VA in 12/14 cases. Transmural discordant alternans was present directly before the induction of VA in three of 14 cases. There were no significant difference in the magnitude, extent or heterogeneity of alternans between hearts which did and did not develop VA.
Conclusions These data demonstrate an increased vulnerability to repolarisation alternans and VA in heart failure, and so establish the experimental paradigm in a clinically relevant pathology. However, there was no demonstrable mechanistic link between transmural alternans and VA in failing hearts. A better understanding of the relationship between alternans and VA will inform the interpretation of clinical data regarding the predictive value of MTWA for VA in patients with heart failure.