Letters Section Editor: Jody W. Zylke, MD, Senior Editor.
Author Affiliations: Rambam Health Care Campus, Haifa, Israel (firstname.lastname@example.org).
To the Editor: The study by Dr Min and colleagues1 offers the possibility of estimating FFR based on CT anatomy alone, with the use of sophisticated computational fluid dynamic techniques. However, several limitations and biases need to be acknowledged.
A significant cause of bias may result from the blinded integration of FFR and CT. The location that corresponded to the point where the actual FFR was measured “was communicated to the FFRCT core laboratory by an arrow on a 3-dimensional volume-rendered CT image of the coronary arteries.” This seems to tell the FFRCT core laboratory where a lesion was suspected on invasive coronary angiography, and thereby reveals where significant lesions were not suspected. This is especially important for calcified lesions and image artifacts, in which it is often difficult to exclude significant disease by CT. The CT core laboratory did not have this knowledge when reporting their findings. It is not clear why, in a patient-based or vessel-based analysis, this integration is necessary. In a real-life clinical scenario, the FFRCT technique would need to work without any prior knowledge of lesion location, other than that available from CT.
Lessick J, Aronson D. Noninvasive Approach to Assess Coronary Artery Stenoses and Ischemia. JAMA. 2013;309(3):233-236. doi:10.1001/jama.2012.157195