II. Various Types of Fusion Between T and U 'Waves

T HE study of the U wave and of its clinical significance depends upon its accurate recognition. Accurate identification of the U wave is also needed for a correct determination of the Q-T duration. This identification is sometimes difficult because of superposition and fusion of the T and U waves. The term "superposition" is applied to patterns in which these waves are only partially merged and a notch is present by which the 2 components can be separated. The term "fusion" implies uniform monophasic or diphasic waves that show no landmarks for the differentiation of their components. Although the partially merged T and U waves have also frequently been misinterpreted as T waves, the greatest difficulty lies in the recognition of TU-fusion patterns. The appearance of TU-fusion waves depends on certain peculiarities of the component waves, which are not the same for monophasic and diphasic patterns. The conditions under which upright TU-fusion waves appear will be investigated first. On analysis of such patterns, 3 types of monophasic TUfusion waves can be differentiated.' Figure 1A demonstrates sinus arrhythmia after mild exercise in a 29-year-old man with neurocirculatory dystonia. At the beginning and at the end of the tracing the U waves are rather high, their summits are drawn near the T waves (T-aU = 0.05 sec.), and the junctions between T and U are elevated. These are typical superposition TU waves. In the middle part of the tracing, on the other hand, during a phase of bradyeardia, there is a uniform broad and high wave in lead II. The apex of this wave is situated between the summits of the T and U waves in lead I, where it can be distinguished well. To understand the mechanism of fusion of the 2 waves in lead II, we must consider that the apex of the T wave occurs later due to the slowing of the rate, that the U wave becomes higher for the same reason, and that the apex of the U wave remains fixed or is even drawn nearer to the T wave. Consideration of lead I shows that both T and U are involved in the formation of the fusion wave in lead II. In figure 1B the 2 forms of merging of T and U in this case are analyzed. For the double TU wave (complex 2 of lead II) we must assume that the U wave begins before the T wave ends but after the apex of T. The fusion TU wave (complex 4 of lead II), on the other hand, is determined by the U wave beginning not only before the end of the T wave, but also at or before the apex of T. This fusion pattern with a new and higher summit may be called a "TUfusion wave with a summation apex" or with "equidominant T and U components," in which the T and U components hardly differ in their amplitude and the U-wave upstroke is steeper than the T-wave downstroke. Figure 2A shows a normal resting electrocardiogram of a 44-year-old man with silicosis. The Q-T duration, the interval between Q and the second heart sound, and the relation between the summits of U and T are normal. Immediately after ascending 6 flights of stairs (fig. 2B) the Q-T interval in the leads I-III and the interval from Q to the second heart sound are shortened more than would be expected at the resulting heart rate. In the precordial leads, however, a higher upright wave appears, the termination of which produces an interval about as long as the expected Q-U interval. Therefore we must assume that this wave is a TU-fusion wave. This mechanism becomes obvious in succeeding tracings. One half minute later (B2) a notch appears that coincides with the notch separating the U wave from the T wave in the limb leads. The emerging U wave is much lower than the T wave. The interval T-aU as well as the interval Q-U appear to be shortened. Figure 3 from the same case shows in dotted lines the resting pattern in leads II and V3, with the second heart sound designated by an arrow, and in solid lines the findings after exercise. The tall waves in V3 cannot be ex-