Feasibility and diagnostic accuracy of exercise treadmill nitrogen-13 ammonia PET myocardial perfusion imaging of obese patients

Treadmill exercise nitrogen-13 (13N)-ammonia positron emission tomography (PET) has logistical challenges and limited literature. We aimed to assess its feasibility, image quality, and diagnostic accuracy in obese and nonobese patients. Between 2009 and 2012, 10,804 patients were referred for myocardial perfusion imaging, including 300 for treadmill PET, of whom 265 were included in this study. Treadmill testing and PET were performed using standard procedures. Image quality, perfusion, and summed stress score (SSS) were assessed. Invasive coronary angiography was performed within 90 days of PET in 43 patients. Mean ± SD body mass index (BMI) was 35.7 ± 7.7 kg/m2 (range 19.5-63.5 kg/m2). Feasibility of treadmill 13N-ammonia PET was 100%. Exercise duration was less for obese patients than nonobese patients (P < .001). Image quality was rated good for 96.9% of obese and 100% of nonobese patients. For all patients, sensitivity was 86.4% and specificity was 74.4%. Diagnostic accuracy did not change significantly with increasing BMI. SSS remained significant in predicting angiographic coronary artery disease after adjustment for age, sex, and Duke treadmill score. Treadmill 13N-ammonia PET is highly feasible, yields good image quality, and has moderately high diagnostic accuracy in a small subset of obese and nonobese patients who are deemed able to perform treadmill exercise.