High Resolution FDG-PET and Coregistered MRI of Atherosclerotic Plaque from a Patient Undergoing Carotid Endarterectomy

In this IRB-approved investigation an 82-year-old male underwent carotid endarterectomy (CEA) for a >80% asymptomatic carotid stenosis. Six weeks prior, the subject had undergone CEA of the contralateral carotid artery for a symptomatic >70% stenosis. Approximately 1 hour prior to surgery, the patient was injected with 7.9 mCi [18]Fluorodeoxyglucose (FDG). After resection of the plaque using a surgical technique that maintained its integrity, the specimen was imaged using micro-PET (Siemens Inveon) and MRI (3T Philips Achieva). Subsequently, the specimen was fixed in formalin, sectioned at 1mm intervals and stained with Hematoxylin and Eosin. Immunocytochemistry was performed using antibodies to macrophages (Ham-56), leukocytes (CD-45) and smooth muscle cell actin (anti - alpha-actin).

The Figure shows matched cross sections from MRI, FDG-PET, and histology obtained at the carotid artery bifurcation and internal carotid artery. The FDG-PET images were first registered and fused with the MRI results, then reformatted to match the corresponding histology sections. FDG uptake is seen to be highly variable with focal hot spots. The maximal standardized uptake value (SUV) for the entire specimen was 3.50 g/cc and occurred in the internal carotid artery (Figure, right). In this zone of high uptake, histology indicated the presence of a necrotic core with recent intraplaque hemorrhage and significant macrophage infiltration, particularly around the periphery of the intraplaque hemorrhage. Other zones of high uptake (Figure, left) were associated with deposition of loose extracellular matrix with concomitant neovasculature and inflammatory infiltrate including macrophages and leukocytes.

Figure.

Ex vivo appearance of carotid atherosclerosis after pre-surgical injection of [18]fluorodeoxyglucose in an 82-year-old male. A maximum intensity projection (MIP) image of the PET volume is shown in the center, with arrows indicating zones of high uptake. PET, MRI, fused PET/MRI and histology (hematoxylin and Eosin [H&E] and HAM-56 at 100X) transverse cross sections are shown for the lower arrow at left and for the upper arrow at right. High PET signals are encoded as yellow, which show spatial correspondence with dense accumulations of macrophages by HAM-56 immunocytochemistry.

To investigate the common features of regions with high SUV, 12 cross sections were analyzed at 2 mm intervals with matched histology. Sixteen regions with SUV in excess of 2.0 were identified by thresholding and the histological features of all 16 regions were recorded. Ten regions (63%) exhibited substantial inflammatory infiltrate including more than 30 macrophages per high-power (600X) field. Five regions (31%) exhibited extensive neovasculature with more than 6 vessels per high-power field. All five of these regions also exhibited loose extracellular matrix. Accumulations of smooth muscle cells were observed in only 2 regions. Necrotic cores with intraplaque hemorrhage were observed in only 2 regions on consecutive slices, which also contained loose matrix and significant numbers of macrophages and neovessels. None of the regions contained calcifications.

Strong uptake of FDG associated with atherosclerosis has previously been reported and associated with extensive inflammatory activity [1,2]. The low in vivo resolution of FDG-PET, however, has precluded exact colocalization of the FDG signal with specific pathology. In fact, some studies suggest the signal may originate outside of the vessel wall [3]. One study used microPET to investigate CEA specimens bathed in [18]FDG [4], but this may not accurately reflect uptake of a perfused plaque in vivo. By injecting [18]FDG prior to surgery, this study was able to definitively localize the FDG signal relative to MRI and histology.