Mechanisms of Microvolt T-Wave Alternans Testing

Otto Constantini, MD
Otto Constantini, MD
I have the difficult task to talk about t-wave alternans on a cellular level to a crew of clinical electrophysiologists. I will try to make it simple. This is a developing story, and it started back in the early 1900s, when visible alternans was described on an EKG on an every-other-beat level in the t-wave (negative, positive, negative, positive, negative, positive...). In this EKG rhythm strip, the alternation is followed by a PVC and then ventricular fibrillation. This is clearly macroscopic t-wave alternans which can be seen on the surface EKG. In later years, it was described in several disease conditions. These strips show patients with congenital long QT syndrome you can see the t-wave alternating after emotional excitement. This EKG shows a patient of ours from the CCU with a dilated cardiomyopathy who also had an episode of t-wave alternans with a negative-positive-negative pattern followed by ventricular fibrillation in the setting of a decompensation of congestive heart failure. Thus, these are macroscopic examples of t-wave alternans. We don't see this very often. The reason is that there are two problems with seeing t-wave alternans on the EKG. The first one is that under controlled conditions in a normal heart, this is clearly a heart-rate dependent phenomenon, and it occurs at fast heart rates, so that at baseline you will not see it. In diseased myocardium, though, the curve is shifted to the left. Patients with structural heart disease develop alternans at lower heart rates, but it is still a rate-dependent phenomenon. This was noticed in the early 1900s, when Thomas Lewis described this as occurring either when the heart muscle is normal but the heart rate is very fast, or when there is serious cardiac disease and the rate is normal. The second reason alternans is not always visible is that the level of alternans on the surface EKG is minuscule compared to the alternans in the action potential at the cellular level, so that when you have 40 millivolts of alternans at the cellular level, you actually have only 0.2 millivolts of alternans on the surface EKG. Therefore, it is very hard to see this on a surface. This is the reason why the stress test t-wave alternans test that we have in place today utilizes special sensitive electrodes and orthogonal leads to pick up the small alternans voltages transmitted from the cell to the surface. The mechanisms have been described using a guinea pig model of pacing-induced t-wave alternans, which uses a wedge preparation of the epicardium and optical mapping techniques in a paced wedge of a guinea pig. To get back to the cellular basis of t-wave alternans, here is another way of describing this. When one paces at a certain frequency, you have no alternans. As you pace faster and faster you develop alternans, but at the EKG level the alternans is very small, in the order of 0.4-0.8 millivolts. Therefore, when you pace at a faster rate you develop alternans; however, the action potential alternans is much higher in magnitude than the EKG alternans. The other thing to note is that pacing at ~275 beats per minute vs. ~315 beats per minute makes a difference. What was noticed is that when you pace at a certain frequency (273 beats per minute), the cells at Point A and Point B alternate with the same frequency in concordance with each other, with Beat #1 in yellow and Beat #2 in green. Therefore, for each beat, the cells alternate in concordance: a long action potential in both A and B and a short action potential with the next beat in both A and B. As you start pacing faster, the cells start changing, and by the time you pace at 316 beats per minute, you have Cell A, in yellow, alternating in a long-short sequence, and Cell B, in green, alternating in a short-long sequence. It is at this point that discordant alternans develops and that dispersion of repolarization occurs, facilitating slowing of conduction in tissue that predisposes one to reentrant mechanisms of arrhythmia. So as you develop discordant alternans, the dispersion of repolarization increases dramatically, to the point where reentrant arrhythmias like ventricular fibrillation and ventricular tachycardia are possible. Here is another example of concordant alternans, with a focus on dispersion of repolarization; this is cell1 and cell2. The cells are alternating in the same pattern. At faster pacing rates, the cells develop discordant alternans. Now, during repolarization, there is first of all slowing of conduction, and secondly the cell is alternating in the opposite direction (white to black, black to white) with every other beat. What does that mean? In a normal preparation, you have depolarization of a paced beat, and you have repolarization of that beat with slowing of conduction. You then have a ventricular premature beat that causes conduction block due to tissue refractoriness, and finally the wave reenters the tissue, causing ventricular fibrillation. What other questions need to be answered? First of all, what is the mechanism responsible for repolarization alternans of individual myocytes? Secondly, what is the mechanism responsible for discordant alternans between myocytes? Two theories have been proposed to explain these phenomena, and though we will not talk about the restitution hypothesis today, I will show you evidence suggesting that calcium handling of the cell plays a large role. We know that in the failing myocardium, calcium handling by the cells is altered. For example, if you look at action potential durations in the normal heart, this slide shows a normal action potential and normal calcium current. In contrast, in congestive heart failure, you have a prolonged action potential, a prolonged QT duration and decreased calcium current. Therefore, calcium handling became a possible mechanism for alternans by the fact that the cell is not able to meet the kinetics of calcium cycling, especially at faster heart rates. The cell has to uptake calcium within it and then reuptake it into the sarcoplasmic reticulum. It then has to release it from the ryanodine receptor, and finally release it from the cell through the sodium exchanger. At certain heart rates, the cell is not able to do this. In these slides, I show you very recent data from Dr. Rosenbaum s group, which shows that as you pace faster and faster (150, 375, 461 beats per minute), you first have no alternans, but then develop minor alternans (127, 135, 127 on the action potential) as well as alternans on the calcium transient (83, 129, 88). At faster heart rates, that difference becomes even more pronounced, and you clearly see alternans on the EKG. The question, however, is what comes first: the action potential alternans or the calcium transient alternans? Through voltage-clamped experiments, you can see that if you have a cell that is unclamped, you have voltage alternans with every other beat and you have calcium alternans (high, low, high, low) with every other beat. If you keep the voltage constant, the calcium continues to alternate, suggesting that calcium is actually the mechanism driving the alternans. Thus, calcium handling is a prime suspect in individual myocyte repolarization alternans. Why then do neighboring cells alternate between each other, and why do they alternate discordantly? There are a couple of hypotheses here, although this is mostly theory and needs to be worked up further. A role may be played by spatial differences in the EP properties of cells across the myocardium; also intercellular uncoupling, such as in patients with myocardial infarction who have a scar or patients with dilated cardiomyopathy who have fibrosis, may play a role. In this slide, one can see these physiological differences in cellular behavior: 1) this is an EKG with a pacing rate of 375 beats per minute which shows alternans with every other beat; 2) this is an alternans-prone site that is alternating at this rate; and 3) this is a site in the preparation that is not yet alternating. At a faster pacing rate, in this slide, all sites alternate further. What that tells us is that there are sites in the preparations that alternate better than others because of the different cellular electrophysiological properties that they have. Some of the biological modifiers of these differences are proteins such as the ryanodine receptor gene, the SERCA gene, or the sodium exchanger gene. They may be implicated in creating electrophysiological differences on the cellular level between areas of the myocardium. You can see in this slide the different expressions between endocardium and epicardium of these proteins. Lastly, the other mechanism that may be implicated in discordant alternans between cells is cellular uncoupling; if you have a preparation in a normal heart with Cell A and Cell B, at a certain pacing rate, they are not alternating. If you develop a structural lesion here we can do this easily in the lab using an epicardial wedge, but this also happens in a myocardial infarction or in fibrosis you can see that at the same pacing rate, the cells now alternate in opposite directions with discordant alternans. So in this slide, in the presence of a scar, and with ventricular pacing, the cells develop discordant alternation, heterogeneities or dispersion of repolarization, and block, predisposing the preparation to ventricular tachycardia. This may be one of the reasons why, for example, ischemic cardiomyopathies may develop ventricular tachycardia, versus a propensity for dilated non-ischemic cardiomyopathies to develop ventricular fibrillation. In conclusion, what is the hypothesis? There are some physiological heterogeneities in the way cells process calcium. Once the heart rate exceeds those kinetics, cellular alternans develop, first concordantly and then discordantly. That transition from concordant to discordant alternans is hastened by ionic channel heterogeneities, by inter-cellular uncoupling, and by some conduction dynamics. All of these yield pathophysiologic heterogeneities which may lead to ventricular fibrillation, polymorphic ventricular tachycardia or monomorphic ventricular tachycardia, depending on whether the patient has a normal structure or structural failures. This story is still developing we are studying why exactly this transition happens and how we get from here to here but it is making a lot more sense, and clearly a lot more work needs to be done in the future. One word of congratulations to all the people who are in this slide, and especially to Dr. Rosenbaum, who has done a vast majority the work that I have just described.