Indeterminate TWA Tests Are Reduced By Ventricular versus Atrial Pacing
Category:  09 Signal Average ECG/T-Wave Alternans
Presentation Time: Friday, 3:45 p.m. - 4:45 p.m.
Scott Vasconcellos, MD, Daejoon Anh, MD, Uma Srivatsa, MD, Nancy McCormick, RN, CNP and Sanjiv M. Narayan, MD, PhD. UCSD and VA Medical Centers, San Diego, CA, Henry Ford Hospital, Detroit, MI
Presentation Number: P5-26
Poster Board Number: P5-26
T-wave alternans (TWA) is a promising risk stratifier for sudden death, yet one-fifth of tests are indeterminate due to inadequate heart rate, excessive ectopy or ECG noise. Indeterminate tests are often considered high-risk, yet this may be less appropriate if they predominantly reflect ECG noise. We hypothesized that indeterminate TWA tests commonly reflect ECG noise, and that their causes would vary with the method of inducing TWA.
Methods and Results. We studied 44 patients with implantable cardioverters (27 biventricular, BV), of age 70+9.6 years, with LVEF 31+10 %, and coronary disease (80%). TWA was measured by pacing from RA, right (RV) and left (LV) ventricle then BV, during acceleration (100-120 beats/min) then deceleration (120-100 beats/min) using HeartWave (Cambridge Heart, MA). Patients were followed for 280+177 days. TWA reports were classified by consensus from 3 physician readers. Automated computer interpretation, in comparison, was 71 % accurate. Of 186 TWA recordings, 21% were negative, 63% were positive and 16% were indeterminate. Inadequate heart rate accounted for 25% of all indeterminate tests, and was significantly higher for RA pacing (16%), reflecting atrioventricular nodal block, than ventricular pacing (0%; p<0.001). Other indeterminate tests reflected noise (68%), excessive bad beats (39%), RR alternans (11%) and other factors (14%). BV pacing trended to reduce bad beats (p=0.1), and all indeterminate tests (10 %) compared to RV (18 %), LV (18 %) and RA (23 %) pacing (p=0.1). Notably, TWA magnitude was higher on deceleration than acceleration at each pacing site (p<0.05), and the symmetrical pacing protocol reduced the number of indeterminates to 16 % (vs 21 % during acceleration). At repeat testing 280+177 days later, TWA reproducibility was 72-76% with ventricular pacing and 67% with RA pacing.
Conclusions. Ventricular pacing is a reliable means of inducing TWA that ensures adequate heart rate, unlike RA pacing, and is reproducible. Notably, the large number of indeterminate tests due to ECG noise suggest caution in grouping them with positive tests (“C rules”).