Cardiac PET-CT Reveals 40% Remedial False-Positive Rate

HOUSTON, July 5 -- Standard helical PET-CT cardiac imaging produces false-positives 40% of the time, but most of the errors can be identified and corrected before inappropriate treatment.
So reported K. Lance Gould, M.D., of the University of Texas here, and colleagues in the July issue of the Journal of Nuclear Medicine, who said that the dual modality continues to expand despite the recognized problems with artifacts and false-positives from misregistration of PET and CT attenuation correction data.
Although fused images are corrected by cine CT PET protocols or commercial software, investigators said, the frequency and the degree to which the defects could be corrected remained unknown prior to this study.
Of 259 consecutive patients who had diagnostic PET-CT perfusion imaging, misregistration of helical PET-CT images led to artifactual PET defects in 103 (40%). The defects were moderate or severe in 59 cases but corrected in each instance by cine or shifted cine PET-CT imaging.
Quantitative misregistration was a strong predictor of artifact size and severity, representing a "real" defect.
Dr. Gould and colleagues prospectively examined images and data from patients undergoing diagnostic dipyradimole perfusion imaging with PET-CT. For each study they used 82Rb, a 16-slice PET-CT scanner, helical CT attenuation correction with breathing and end-expiratory breath-hold, and averaged cine CT data during breathing.
The researchers objectively measured misregistration of PET-CT fusion images in millimeters and correlated the findings with the size and severity of the associated defect. They defined misregistration artifacts as PET defects with corresponding misregistration on helical PET-CT fusion images that resolved after correct coregistration. They established this by using a repeat CT scan, cine CT averaged attenuation during normal breathing, or shifted cine CT data that coregistered with PET data.
"[T]o be counted as having an abnormality due to PET/CT misregistration, every patient with abnormal helical CT-PET at rest or stress had to have a stress perfusion study with no significant regional defect after correct PET-CT coregistration," the investigators wrote. "Any patient with a defect that persisted on stress PET/CT images or after correct coregistration that was outside [two standard deviations] of healthy subjects was classified as having a "real" defect. . . ."
In 137 of the 259 cases, patients were classified as having real perfusion abnormalities that persisted after correction of PT-CT misregistration or on stress PET. The remaining 122 cases comprised imaging studies with and without misregistration artifacts. Quantitative analysis identified additional patients with true perfusion defects outside two standard deviations of healthy controls.
Multivariate analysis was performed to gain insight into the factors precipitating the false results. Transaxial misregistration (x-axis) proved to be the strongest predictor of artifact size and severity. Transaxial misregistration of more than 6 mm in anterior or lateral areas occurred in 76% of the artifactual defects, in inferior areas in 16%, and at the apex in 8%.
The authors acknowledged that the lack of coronary arteriography to document the absence of stenosis could be viewed as a limitation of their study. However, they argued that arteriography is not justified for a defect in a region of PET-CT misregistration that disappears after reconstruction using correctly coregistered CT attenuation data.
Dr. Gould and colleagues concluded that properly coregistered PET-CT data provide definitive perfusion images that are suitable for assessing the severity of coronary artery disease and for identifying early nonobstructive disease that could be targeted by intension pharmacologic therapy and lifestyle management.
The authors reported no conflicts of interest related to this study.Primary source: Journal of Nuclear MedicineSource reference: Gould KL et al. "Frequent diagnostic errors in cardiac PET/CT due to misregistration of CT attenuation and emission PET images: a definitive analysis of causes, consequences, and corrections." J Nucl Med 2007;48:1112-1121.