Quantifying myocardial flow reserve with PET aids in CAD diagnosis
nuclear cardiology - 94.12 Kb
Normal PET myocardial perfusion imaging (MPI) in coronary artery disease. Cardiac 13N-ammonia PET shows normal MPI at rest and at adenosine-stress but abnormal myocardial flow reserve, indicating global myocardial under perfusion. Source: J Nucl Med 2012;53(8):1230-1234
The quantification of myocardial flow reserve (MFR) in 13N-ammonia PET/CT myocardial perfusion imaging (MPI) provides a substantial added diagnostic value for detection of coronary artery disease (CAD), according to study results in the August issue of the Journal of Nuclear Medicine. In patients with normal MPI results, quantification of MFR helps to unmask clinically significant CAD.

The capability to obtain quantitative values of flow and MFR has been perceived as an important advantage of PET over conventional nuclear MPI. “Although many studies have revealed a reversed correlation of increasing coronary artery lesion narrowing with decreasing hyperemic flow and MFR in the respective myocardial territory, its diagnostic added value over MPI PET has not been assessed systematically,” the authors wrote.

For this study, Michael Fietcher, MD, of the department of radiology at the University Hospital Zurich in Switzerland, and colleagues evaluated the hypothesis that patients as assessed with 13N-ammonia and PET/CT with decreased MFR (less than two) would have a higher probability of CAD and thus, MFR would confer an added diagnostic value over MPI alone to predict angiographically documented CAD (coronary luminal narrowing at least 50 percent).

Seventy-three patients underwent a one-day adenosine stress-rest with 13N-ammonia PET/CT MPI. The researchers calculated the MFR and also evaluated the added value of MFR as an adjunct to MPI for predicting CAD (luminal narrowing at least 50 percent) using invasive coronary angiography as a standard of reference.

Per patient, the sensitivity, specificity, positive predictive value, negative predictive value and accuracy of MPI for detecting significant CAD were 79 percent, 80 percent, 91 percent, 59 percent and 79 percent, respectively. Adding a cutoff of less than two for global MFR to MPI findings improved the values to 96 percent, 80 percent, 93 percent, 89 percent and 92 percent, respectively.

Adding MFR to MPI reclassified a third of all normal MPI findings, wrote Fietcher et al. As a result, accuracy increased significantly after the addition of MFR from 79 percent to 92 percent, mainly because of an increase in sensitivity from 79 percent to 96 percent, that is, from the lower end to clearly above the range of sensitivities reported in a recent meta-analysis (Acad Radiol 2008;15:444–451).

The fact that CAD prevalence was 73 percent “may be perceived as a potential limitation” of the study, according to the authors, as such a high prevalence usually tends to yield a lower negative predictive value. This notion would also hold true after adding the MFR information, they added. Furthermore, the present study offers diagnostic data but was not designed to report prognostic information.

In defense of their findings, Fietcher et al said that the fact that negative predictive value increased substantially from 59 percent to 89 percent after MFR was implemented “underlines its strong clinical validity.”

They also recommended that future studies should be designed to evaluate the impact of a PET-guided choice of treatment strategy on outcome.