Comparison of C-11-meta-hydroxyephedrine and N-13-ammonia PET Imaging Biomarkers of Cardiac Sympathetic Innervation and Perfusion in Ischemic Cardiomyopathy

Background: Dysfunction of the cardiac sympathetic nervous system is a predictor of sudden cardiac arrest in patients with ischemic cardiomyopathy. Results of the PAREPET trial suggested that markers of sympathetic nervous system (SNS) function may be more sensitive than myocardial perfusion in detecting regions of arrhythmogenic myocardium in these patients. However, it is unknown to what extent C-11-meta-hydroxyephedrine, a marker of cardiac SNS integrity, provides diagnostic information independent of blood flow given that it requires adequate tracer delivery to the myocardium via tissue perfusion. Objective: The objective of this study was to compare the extent and severity of abnormalities in left ventricle (LV) myocardial perfusion and sympathetic innervation measured in the PAREPET trial, using N-13-ammonia (NH3) and C-11-meta-hydroxyephedrine (HED) respectively. This trial included patients with documented ischemic cardiomyopathy (LVEF&#8804; 35%) and primary-prevention ICDs. Methods: Two reviewers rated the quality of all HED and NH3 dynamic PET scans on a 4-point scale (very good, good poor, non-diagnostic) based upon factors including image contrast and noise, as well as adequate spatial and temporal sampling. An automated analysis program was used to measure regional defect scores (%LV extent × severity <75% of the maximum value) on polar-maps of HED and NH3 ‘qualitative’ uptake activity, as well as ‘quantitative’ HED distribution volume (DV) and NH3 myocardial blood flow (MBF). To evaluate the flow-dependence of HED uptake and clearance, the %LV defect scores were compared between HED and NH3 in tertiles of perfusion and using linear regression.

Results: N=135 HED and NH3 scan pairs with good or very good image quality were identified. The %LV defect scores measured by qualitative HED tracer uptake were on average 10% ± 8% higher than those measured by NH3 uptake (p<0.001). This difference was similar when comparing the quantitative HED DV scores to NH3 MBF defect scores (+10% ± 13% p<0.001), and was even larger when comparing HED DV to the qualitative NH3 uptake scores (+19% ± 10%; p<0.001). The largest differences in HED vs NH3 defect scores (+12% to +21%) were observed in the bottom tertile of perfusion. These differences decreased progressively with increasing perfusion, resulting in similar values in the top tertile for quantitative HED DV vs NH3 MBF defect scores (43% vs 39%; p=NS). There was moderate correlation (R²=0.43) between the qualitative HED and NH3 uptake defect scores, that decreased to R²=0.30, when comparing quantitative HED DV vs qualitative NH3 scores. The lowest correlation (R²=0.18) was found between the quantitative HED DV vs NH3 MBF defect scores, consistent with the highest intercept and lowest slope (y = 27+0.36x %LV) reflecting minimal dependence of the quantitative HED DV scores on MBF.

Conclusions: Regional abnormalities of HED uptake and distribution volume were larger and/or more severe compared to NH3 uptake and MBF respectively, in the LV myocardium of patients with ischemic cardiomyopathy. The largest differences were observed in the patients with relatively mild perfusion defects. Our results confirm previous findings that the extent and severity of sympathetic denervation generally exceeds that of hypo-perfusion. They support the incremental value of SNS over MBF PET imaging in this patient population, and suggest there may be further value in absolute quantification of HED kinetics over qualitative imaging of tracer uptake alone.