Search

The GE Discovery NM (DNM) 530c/570c are dedicated cardiac SPECT scanners with 19 detector modules designed for stationary imaging. This study aims to incorporate additional projection angular sampling to improve reconstruction quality. A deep learning method is also proposed to generate synthetic dense-view image volumes from few-view counterparts.By moving the detector array, a total of four projection angle sets were acquired and combined for image reconstructions. A deep neural network is proposed to generate synthetic four-angle images with 76 ([Formula: see text]) projections from corresponding one-angle images with 19 projections. Simulated data, pig, physical phantom, and human studies were used for network training and evaluation. Reconstruction results were quantitatively evaluated using representative image metrics. The myocardial perfusion defect size of different subjects was quantified using an FDA-cleared clinical software.Multi-angle reconstructions and network results have higher image resolution, improved uniformity on normal myocardium, more accurate defect quantification, and superior quantitative values on all the testing data. As validated against cardiac catheterization and diagnostic results, deep learning results showed improved image quality with better defect contrast on human studies.Increasing angular sampling can substantially improve image quality on DNM, and deep learning can be implemented to improve reconstruction quality in case of stationary imaging.

Synapse loss is a hallmark of Alzheimer’s disease (AD) and correlates with cognitive decline. The validation of a noninvasive in vivo imaging approach to quantify synapse would greatly facilitate our understanding of AD pathogenesis and assist drug developments for AD. As animal models of neurodegenerative and neuropsychiatric disorders play a critical role in the drug discovery and development process, a robust, objective, and translational method for quantifying therapeutic drug efficacy in animal models will facilitate the drug development process. In this study, we tested the quantification reliability of the SV2A PET tracer, [18F]SynVesT-1, in a mouse model of AD (APP/PS1) and wild-type controls, and developed a simplified quantification method to facilitate large cohort preclinical imaging studies. We generated nondisplaceable binding potential (BPND) and distribution volume ratio (DVR) values using the simplified reference tissue model (SRTM) on the 90-min dynamic PET imaging data, with brain stem and cerebellum as the reference region, respectively. Then, we correlated the standardized uptake value ratio (SUVR)-1 and SUVR averaged from different imaging windows with BPND and DVR, using brain stem and cerebellum as the reference region, respectively. We performed homologous competitive binding assay and autoradiographic saturation binding assay using [18F]SynVesT-1 to calculate the Bmax and Kd. Using brain stem as the reference region, the averaged SUVR-1 from 30 to 60 min postinjection correlated well with the BPND calculated using SRTM. Using cerebellum as the reference region, the averaged SUVR from 30 to 60 min postinjection correlated well with the SRTM DVR. From the homologous competitive binding assay and autoradiographic saturation binding assay, the calculated the Bmax and Kd were 4.5–18 pmol/mg protein and 9.8–19.6 nM, respectively, for rodent brain tissue. This simplified SUVR method provides reasonable SV2A measures in APP/PS1 mice and their littermate controls. Our data indicate that, in lieu of a full 90-min dynamic scan, a 30-min static PET scan (from 30 to 60 min postinjection) would be sufficient to provide quantification data on SV2A expression, equivalent to the data generated from kinetic modeling. The methods developed here are readily applicable to the evaluation of therapeutic effects of novel drugs in this rodent model using [18F]SynVesT-1 and small animal PET.

Amivantamab is a novel bispecific antibody that simultaneously targets the epidermal growth factor receptor (EGFR) and the hepatocyte growth factor receptor (HGFR/c-MET) that are overexpressed in several types of cancer including triple-negative breast cancer (TNBC). Targeting both receptors simultaneously can overcome resistance to mono-targeted therapy. The purpose of this study is to develop 89Zr-labeled amivantamab as a potential companion diagnostic imaging agent to amivantamab therapy using various preclinical models of TNBC for evaluation. Amivantamab was conjugated to desferrioxamine (DFO) and radiolabeled with 89Zr to obtain [89Zr]ZrDFO-amivantamab. Binding of the bispecific [89Zr]ZrDFO-amivantamab as well as its mono-specific “single-arm” antibody controls were determined in vitro and in vivo. Biodistribution studies of [89Zr]ZrDFO-amivantamab were performed in MDA-MB-468 xenografts to determine the optimal imaging time point. PET/CT imaging with [89Zr]ZrDFO-amivantamab or its isotype control was performed in a panel of TNBC xenografts with varying levels of EGFR and c-MET expression. [89Zr]ZrDFO-amivantamab was synthesized with a specific activity of 148 MBq/mg and radiochemical yield of ≥ 95%. Radioligand binding studies and western blot confirmed the order of EGFR and c-MET expression levels: HCC827 lung cancer cell (positive control) > MDA-MB-468 > MDA-MB-231 > MDA-MB-453. [89Zr]ZrDFO-amivantamab demonstrated bispecific binding in cell lines co-expressed with EGFR and c-MET. PET/CT imaging with [89Zr]ZrDFO-amivantamab in TNBC xenografted mice showed standard uptake value (SUVmean) of 6.0 ± 1.1 in MDA-MB-468, 4.2 ± 1.4 in MDA-MB-231, and 1.5 ± 1.4 in MDA-MB-453 tumors, which are consistent with their receptors’ expression levels on the cell surface. We have successfully prepared a radiolabeled bispecific antibody, [89Zr]ZrDFO-amivantamab, and evaluated its pharmacologic and imaging properties in comparison with its single-arm antibodies and non-specific isotype controls. [89Zr]ZrDFO-amivantamab demonstrated the greatest uptake in tumors co-expressing EGFR and c-MET.