Author + information
- Morten Sengeløv, MB∗ (, )
- Peter Godsk Jørgensen, MD,
- Jan Skov Jensen, MD, PhD, DMSc, MD,
- Niels Eske Bruun, MD, DMSc,
- Flemming Javier Olsen, MB,
- Thomas Fritz-Hansen, MD,
- Kotaro Nochioka, MD, PhD and
- Tor Biering-Sørensen, MD, PhD
- ↵∗Department of Cardiology, Gentofte Hospital, University of Copenhagen, Denmark, Kildegårdsvej 28, DK-2900, Post 835, Copenhagen, Denmark
We thank Dr. Zaïni and colleagues for the interest in our paper (1). We appreciate the importance of standardizing the region of interest (ROI) placement when obtaining global longitudinal strain (GLS). In case of myocardial wall thinning, the segment was excluded if the investigator assessed that it was being inadequately tracked. Therefore, we would recommend clinicians to carefully visually assess each ROI segment and exclude those whose traces are not compatible with speckle tracking when obtaining GLS. Our research group has previously validated the interobserver and intraobserver reproducibility of GLS in an ischemic patient population with good results (2). We are not able to provide test-retest variability because our study is retrospective.
To further comment on this issue, the software (Echopac BT12, GE Vingmed, Horton, Norway) used at the time of analysis did not allow adjustment of the width of individual segments. However, BT12 and GLS are comparable to a different vendor and to the newer version, BT13 (3). BT13 allows for readjustment of the individual segments in order to comply with segmental width differences. It would have been interesting to use this software in the case of asymmetrical wall thickness, but the software was not available at the time of analysis. Though an intriguing concept, the authors are not aware of any study examining the effect of regional ROI width differences, and what implications it has for GLS.
In the multivariable Cox regression, we only included echocardiographic parameters that were significant predictors of mortality in the univariable analysis. Left ventricular internal dimension was not a univariable predictor (Table 2 in our paper ) and was therefore not included. We cannot exclude the possibility that the prognostic superiority of GLS is derived from patients with small hypertrophied ventricles; however, we do not believe this speculation can be justified by using our data alone considering that we included left ventricular mass index in the multivariable model. We believe GLS may be a more sensitive measure of longitudinal systolic performance compared with left ventricular ejection fraction, which essentially is a measure of volumetric change.
We agree that it is important to account for cardiovascular mortality. As stated, this is a retrospective study and in this setting, all-cause mortality was the most unbiased endpoint to retrieve since all deaths are registered by the Danish Civil Registration System, resulting in 100% follow-up. However, we are looking into retrieving secondary endpoints such as cardiovascular mortality and, based on intuition, we think this will further improve the prognostic strength of GLS in our cohort.
Please note: This study was financially supported by the Danish Council for Independent Research. The authors have reported that they have no relationships relevant to the contents of this paper to disclose.
- 2016 American College of Cardiology Foundation
- Sengeløv M.,
- Jørgensen P.G.,
- Jensen J.S.,
- et al.
- Biering-Sorensen T.,
- Hoffmann S.,
- Mogelvang R.,
- et al.