29 Jan 2015

Clinicians can now choose between three approved PET Aβ imaging tracers: Eli Lilly’s florbetapir (Amyvid), GE Healthcare’s flutemetamol (Vizamyl), and Piramal Imaging’s florbetaben (Neuraceq). Are there significant differences in their ability to detect Aβ in the brain? A new analysis suggests not. In the January 26 JAMA Neurology, researchers led by Stephen Salloway at Butler Hospital, Providence, Rhode Island, formally published results from the Phase 3 study of flutemetamol. Among 68 patients who came to autopsy, flutemetamol detected brain amyloid with a median sensitivity and specificity of 88 percent each. David Wolk, University of Pennsylvania in Philadelphia, had previously presented the findings at the 64th annual meeting of the American Academy of Neurology (see May 2012 conference news).

How does this performance stack up to those of the other tracers? The question presents a challenge because the three agents have never been measured in the same trial, and available data from their separate studies indicated that some have higher sensitivity and some higher specificity. To compare the three, Salloway and colleagues examined receiver operating characteristic (ROC) plots of sensitivity. In these graphs, a single measurement, the area under the curve, provided a measure of performance for each tracer. The result was a statistical dead heat. All three came out with similar numbers: 0.90 for flutemetamol and florbetaben, and 0.85 for florbetapir.—Madolyn Bowman Rogers