Author + information
- David Wallace-Bradley, B.Sci and
- Juan F. Granada, MD⁎ ()
- ↵⁎Skirball Center for Cardiovascular Research, The Cardiovascular Research Foundation, 8 Corporate Drive, Orangeburg, New York 10962
We are glad to respond to the letter sent by Ferrante et al. and appreciate the time taken to review and comment on our article (1) comparing the accuracy of optical coherence tomography and histopathology in the assessment of the neointimal response following drug-eluting stent implantation in a pre-clinical model of coronary disease. It is important to highlight that the objective of our study was to compare the findings of data obtained in a clinically relevant fashion to histopathology findings. At first glance, the quantitative differences observed between all measured variables were minimal when both techniques were compared. In our study, regardless of the type of statistical methodology used, our findings suggested a high level of correlation between the 2 methods for the measurement of lumen areas, neointimal areas, and neointimal thicknesses. Although we are grateful for the comments regarding our statistical analysis methodology, we maintain our original position with regards to our paper's conclusions.
In particular, we agree that the inclusion of the fitted correlation line equation would elucidate the validity of our findings even further. Apart from the high R2 values demonstrated between lumen areas, neointimal areas, and neointimal thicknesses measured by optical coherence tomography and histology, their points lie along a line very close to the line of equality in all measured parameters. As was stated, our analysis revealed a limited correlation between methodologies for the measurement of stent area. Furthermore, differences from the origin in the intercept of the lumen and neointimal areas correlation lines are due to slight differences in their measurements found throughout the samples, explanations for which were hypothesized within the paper.
Although Bland-Altman plots for the visualization of the method measurement differences were also constructed, their inclusion in the paper was decided against in the interests of space. As a matter of fact, Bland-Altman plots for all variables analyzed, besides stent areas, did not uncover any differences between methodologies examined in this study. The Bland-Altman plot developed for the evaluation of neointimal thickness is shown. As can be seen, the mean of the differences are very close to 0 for all parameter measures with few outliers. Furthermore, it is important to clarify that the Barlis article referred to by Ferrante et al. and plots therein are primarily concerned with optical coherence tomography interobserver and intraobserver measurement variability and not the contrast of heterogeneous measurement modalities.
- American College of Cardiology Foundation