Author + information
- David Spragg, MD⁎ (, )
- David Kass, MD and
- Ronald Berger, MD, PhD
- ↵⁎Division of Cardiology, Johns Hopkins Hospital, Carnegie 568, 600 North Wolfe Street, Baltimore, Maryland 21287
We welcome the comments from Drs. Wichterle and Vancura regarding our paper (1) and embrace the opportunity to discuss the question further. We agree that the more sites that are sampled and considered as independent observations, the greater the probability is for a difference to appear due to statistical variance alone. This is particularly relevant when only 1 site is measured for the comparator. However, to be the sole explanation, sites with higher responses would be independent findings randomly distributed about the ventricle, yet this was not found to be the case. Rather, we saw the results cluster in geographic territories, both for sites where the dP/dtmax change was low and where it was enhanced. Optimal pacing regions typically were composed of 3 to 6 individual sampled sites, each yielding very similar responses. To pursue the effect of observation multiplicity further, we repeated our analysis but used the average dP/dtmax values from these multiple sampling sites within the entire region (typically 4 sites spread over a 13-cm2 area). Importantly, the optimal regions were still significantly improved over epicardial pacing, although predictably the magnitude of the difference (7.5%) was less than the 13% absolute enhancement from the single best site. Moreover, we found consistent patterns among different patients, further suggesting that these findings were not simply due to chance. In 8 of 11 patients, the left ventricular (LV) base provided the optimal LV mechanical response, whereas such sites were not found in any patient on the inferior wall. This indicates that the results were not driven by randomly distributed statistical outliers.
The primary focus of our study was to assess the size and distribution of an effective zone of endocardial LV stimulation to achieve cardiac resynchronization therapy in heart failure patients with extensive ischemic disease. For this goal, the statistical issues raised do not affect the interpretation. Although we agree that the magnitude of difference that we found between best endocardial and sole epicardial site may have been somewhat influenced by methodology, the data still support the conclusion that in patients with ischemic cardiomyopathy, pacing within a region of the LV endocardium can provide improved mechanical response over clinically selected epicardial pacing sites and that most often, this endocardial region is at the lateral LV base.
- American College of Cardiology Foundation