Diagnostic performance of angiography-derived fractional flow reserve: a systematic review and Bayesian meta-analysis

Aims: Pressure-wire assessment of coronary stenosis is considered the invasive reference standard for detection of ischaemia-generating lesions. Recently, methods to estimate the fractional flow reserve (FFR) from conventional angiography without the use of a pressure wire have been developed, and were shown to have an excellent diagnostic accuracy. The present systematic review and meta-analysis aimed at determining the diagnostic performance of angiography-derived FFR for the diagnosis of haemodynamically significant coronary artery disease. Methods and results: A systematic review and meta-analysis of studies assessing the diagnostic performance of angiography-derived FFR systems were performed. The primary outcome of interest was pooled sensitivity and specificity. Thirteen studies comprising 1842 vessels were included in the final analysis. A Bayesian bivariate meta-analysis yielded a pooled sensitivity of 89% (95% credible interval 83-94%), specificity of 90% (95% credible interval 88-92%), positive likelihood ratio (+LR) of 9.3 (95% credible interval 7.3-11.7) and negative likelihood ratio (-LR) of 0.13 (95% credible interval 0.07-0.2). The summary area under the receiver-operating curve was 0.84 (95% credible interval 0.66-0.94). Meta-regression analysis did not find differences between the methods for pressure-drop calculation (computational fluid dynamics vs. mathematical formula), type of analysis (on-line vs. off-line) or software packages. Conclusion: The accuracy of angiography-derived FFR was good to detect haemodynamically significant lesions with pressure-wire measured FFR as a reference. Computational approaches and software packages did not influence the diagnostic accuracy of angiography-derived FFR. A diagnostic strategy trial with angiography-derived FFR evaluating clinical endpoints is warranted.