Author + information
- Damini Dey,
- Erick Alexanderson Rosas,
- Annika Schuhbaeck,
- Yuka Otaki,
- Luis EJ Orozco,
- Ryo Nakazato,
- Guido Germano,
- Aloha Meave,
- Daniel Berman,
- Stephan Achenbach and
- Piotr Slomka
Coronary CT Angiography (CTA) allows direct, noninvasive evaluation of the coronary arteries. We aimed to evaluate whether integrated quantitative measurement of adverse plaque features from coronary CTA predicts impaired Myocardial Flow Reserve (MFR) and ischemia, measured by 13N-ammonia PET myocardial perfusion imaging.
Thirty-seven patients underwent combined rest-stress 13N-ammonia PET and coronary CTA scans by hybrid PET/CT. Regional MFR and ischemia (% stress-rest perfusion deficit) were automatically derived from PET. From CT, 111 arteries were evaluated by automated software, computing arterial non-calcified (NCP), low-density NCP, and total plaque, as well as maximum diameter stenosis, maximum remodeling index and maximum contrast density drop over the lesions. Visually, maximum stenosis was assessed by standard grading (0-6). Quantitative CT measures were combined by boosted ensemble machine learning algorithm into a composite score to predict impaired MFR (MFR ≤ 2.0) and ischemia (≥2% stress-rest perfusion deficit) by PET, in each artery.
Patients with impaired regional MFR had significantly higher NCP, low density NCP and total plaque in the corresponding arteries (NCP: 156.9 ± 152.9 vs 82.6 ± 124.7 mm3[p=0.02]; low density NCP: 36.8 ± 44.9 vs 19.4 ± 31.9 mm3[p=0.04]; total plaque: 168.6 ± 163.0 vs 93.3 ± 142.7 mm3[p=0.03]). For prediction of impaired MFR, Receiver Operator Characteristic area-under-curve (AUC) for the composite score was 0.81, higher than for visual stenosis grade (0.60, p<0.0001); score influence of contrast density drop was 33.2%, followed by quantitative stenosis (9.4%), remodeling index (9.0%) and NCP (5.0%). For prediction of ischemia, AUC for the composite score was 0.84, higher than for visual stenosis grade (0.60, p<0.0001); score influence of NCP was 14.6%, followed by quantitative stenosis (13.4%), contrast density drop (11.5%) and remodeling index (11.1%).
Integrated quantitative analysis of adverse plaque features from CTA allows significantly improved prediction of impaired MFR and ischemia compared to visual stenosis grade.
Poster Sessions, Expo North
Sunday, March 10, 2013, 3:45 p.m.-4:30 p.m.
Session Title: Imaging: CT/Multimodality VII
Abstract Category: 20. Imaging: CT/Multimodality
Presentation Number: 1269-347
- 2013 American College of Cardiology Foundation