Microvolt-level T-wave alternans determination using the spectral method in patients with QT prolongation: value of adjusting the T-wave window

Abstract

Background: Microvolt-level T-wave alternans (MTWA) measured by the spectral method is a useful risk predictor for sudden cardiac death because of its high negative predictive value. MTWA analysis software selects a segment of the ECG that encompasses the T-wave in most individuals, but may miss the T-wave end in patients with QT prolongation.

Hypotheses: (1) In patients with QT prolongation, adjustment of the T-wave window will increase the sensitivity of MTWA detection. (2) The extent of T-wave window adjustment needed will correspond to the degree of QT prolongation.

Methods: Using data from long-QT syndrome patients, including QTc <0.45 s (normal), 0.45-0.49 s (moderate prolongation), and ≥ 0.50s (severe prolongation), MTWA analysis was performed before and after T-wave window adjustment.

Results: Of 119 patients, 74% required T-wave window adjustment. There was a stronger association between the magnitude of the T-wave offset and the unadjusted QT than between the magnitude of the T-wave offset and QTc (Spearman correlation coefficient 0.690 vs. 0.485 respectively, P<.05). Of 99 initially negative MTWA results, 4 became non-negative after adjustment of the T-wave window (P<.05). All 8 initially positive studies and 12 initially indeterminate studies remained positive and indeterminate, respectively.

Conclusions: T-wave window adjustment can enable detection of abnormal MTWA that otherwise would be classified as "negative" or "normal." Newly developed T-wave window adjustment software may further improve the negative predictive value of MTWA testing and should be validated in a structural heart disease population.

Figure 1. The Automatically-Determined T-wave Window in a Patient with Severe QT Prolongation

Figure 1 shows an example of an automatically positioned T-wave window, marked by the caliper labeled “Tbeg” to the caliper marked “Tend” in a patient with severe QT prolongation (> 500 ms). Note that in this instance, the late portion of the T-wave occurs outside of the T-wave window.

Figure 2. T-Wave Offset versus QT Interval

Figure 2 shows the timing interval by which the T-wave window was advanced (offset), in milliseconds, versus the unadjusted QT interval in seconds.

Figure 3. MTWA Analyzed before and after T-Wave Window Adjustment

Figure 3 shows one of the tests which converted from negative to positive. In the left panel is MTWA analyzed using standard software. At the top is the heart rate profile during exercise and cool-down, with a maximum heart rate of 117/min. From top to bottom, the next tracing shows a low level of “bad” (ectopic) beats, and the next shows an acceptably low level of noise. Below these tracings are the vector magnitude and X, Y and Z leads. Note that in VM and X there is sustained MTWA which begins at an onset heart rate of 111/min as determined by the automatic reader, consistent with a negative study. In the right panel is MTWA reanalyzed after T-wave window adjustment. Note that the automatic reader has established the onset of MTWA at a heart rate of 96/min, consistent with a positive study.

Figure 4. The T-Wave Window during Rest and Exercise, before and after Adjustment

Figure 4 shows the automatically generated (panel A) and manually adjusted (panel B) T-wave windows for the patient whose MTWA results appear in figure 3, (i) at rest, (ii) during exercise at heart rate 97 bpm, and (iii) at peak exercise with heart rate 117 bpm. Note that the T-wave window was manually adjusted at rest (panel B, top) and that the width of this window was then automatically shortened by the software as heart rate increased, so as to avoid including the P wave of the following beat. The asterisk indicates that MTWA was observed during exercise at 97 bpm after adjustment of the T-wave window; MTWA was also present at peak exercise both before and after T-wave window adjustment.