Evaluation of a semi-automated approach for FDG PET image analysis for routine clinical application in patients with multiple myeloma.

BACKGROUND: FDG PET/CT is a tool for assessing response to therapy in various cancers, and may provide an earlier biomarker of clinical response. We developed a novel semi-automated approach for analyzing FDG PET/CT images in patients with multiple myeloma (MM) to standardize FDG PET application. METHODS: Patients (n = 8) with relapsed/refractory MM from the Phase 2 study (NCT02899052) of venetoclax plus carfilzomib and dexamethasone underwent FDG PET/CT at baseline and up to two timepoints during treatment. Images were processed using an established automated segmentation algorithm, with the modification that a red marrow region in an unaffected lumbar vertebra was used to define background standardized uptake value normalized to lean body mass (SUL) threshold above which uptake was considered disease-specific uptake. This approach was compared to lesion segmentation, and to International Myeloma Working Group (IMWG) response criteria, including minimal residual disease (MRD). RESULTS: The two FDG PET analysis techniques agreed on evaluation of patient-level SULpeak for 67% of scans. In the metabolic response assessment per PET Response Criteria in Solid Tumors (PERCIST), the two techniques agreed in 75% of patients. Differences between techniques occurred in low-uptake lesions due to greater reader sensitivity to lesions with uptake marginally above background. PERCIST outcomes were generally in agreement with IMWC and MRD. CONCLUSIONS: This semi-automated analysis was in high agreement with standard approaches for detecting response to MM therapy. This proof-of-concept study suggests that larger studies should be conducted to confirm how FDG PET analysis may aid early response detection in MM.

Intrathecal 99mTc-DTPA imaging of molecular passage from lumbar cerebrospinal fluid to brain and periphery in humans.

INTRODUCTION: Cerebrospinal fluid (CSF) molecular exchange with brain interstitial fluid (ISF) and periphery is implicated in neurological disorders but needs better quantitative clinical assessment approaches. METHODS: Following intrathecal (ITH) dosing via lumbar puncture, Technetium-99 m (99mTc-) diethylenetriaminepentaacetic acid (DTPA) imaging was used to quantify neuraxial spread, CSF-brain molecular exchange, and CSF-peripheral clearance in 15 normal human volunteers. The effect of experimental convection manipulation on these processes was also assessed. RESULTS: Rostral cranial 99mTc-DTPA exposures were influenced by the volume of artificial CSF in the formulation. Signal translocation to the cranial cisterns and the brain parenchyma was observable by 3 hours. 99mTc-DTPA penetrated cortical ISF but showed lower signal in deeper structures. Urinary 99mTc-DTPA signal elimination was accelerated by higher formulation volumes and mechanical convection. DISCUSSION: Widely used for detecting CSF leaks, ITH 99mTc-DTPA imaging can also become a useful clinical biomarker for measuring molecular exchange physiology between the CSF, brain, and periphery.

Dual-Isotope Cryoimaging Quantitative Autoradiography: Investigating Antibody-Drug Conjugate Distribution and Payload Delivery Through Imaging.

In vitro properties of antibody-drug conjugates (ADCs) such as binding, internalization, and cytotoxicity are often well characterized before in vivo studies. Interpretation of in vivo studies might be significantly enhanced by molecular imaging tools. We present here a dual-isotope cryoimaging quantitative autoradiography (CIQA) methodology combined with advanced 3-dimensional imaging and analysis allowing for the simultaneous study of both antibody and payload distribution in tissues of interest in a preclinical setting. Methods: TAK-264, an investigational ADC targeting anti-guanylyl cyclase C (GCC), was synthesized using tritiated monomethyl auristatin E. The tritiated ADC was then conjugated to diethylenetriaminepentaacetic acid, labeled with 111In, and evaluated in vivo in animals bearing GCC-positive and GCC-negative tumors. Results: CIQA revealed the time course of drug release from ADC and its distribution into various tumor regions that are less accessible to the antibody. For GCC-positive tumors, a representative section obtained 96 h after tracer injection showed only 0.8% of the voxels to have colocalized signal, versus over 15% of the voxels for a GCC-negative tumor section, suggesting successful and specific cleaving of the toxin in the GCC-positive lesions. Conclusion: The combination of a veteran established autoradiography technology with advanced image analysis methodologies affords an experimental tool that can support detailed characterization of ADC tumor penetration and pharmacokinetics.

Three-Dimensional Dosimetry for Radiation Safety Estimates from Intrathecal Administration.

Intrathecal administration is of growing interest for drug delivery, and its utility is being increasingly investigated through imaging. In this work, the 3-dimensional Voxel-Based Internal Dosimetry Application (VIDA) and 4D Extended Cardiac Torso Phantom (XCAT) were extended to provide radiation safety estimates specific to intrathecal administration. Methods: The 3-dimensional VIDA dosimetry application Monte Carlo simulation was run using a modified XCAT phantom with additional and edited cerebrospinal fluid (CSF) regions to produce voxel-level absorbed dose per unit cumulated activity maps for 9 selected source regions. Simulation validation was performed to compare absorbed dose estimates for common organs in a preexisting dosimetry tool (OLINDA/EXM). Dynamic planar imaging data were acquired in 6 healthy subjects using administered volumes of 5 or 15 mL (n = 3 each) of 185 MBq of 99mTc-diethylenetriaminepentaacetic acid. Absorbed dose was estimated for each subject using the intrathecal-specific dosimetry application. Results: Simulation results were within 6% of OLINDA estimates for common organs. Absorbed dose estimates were highest (0.3-0.8 mGy/MBq) in the lumbar CSF space. A whole-body effective dose estimate of 0.003 mSv/MBq was observed. An administered volume dependency was observed with a 15-mL volume, resulting in lower absorbed dose estimates for several intrathecal and nonintrathecal regions. Conclusion: The intrathecal-specific VIDA implementation enables tailored dosimetry estimation for regions most relevant in intrathecal administration. Absorbed doses are highly localized to CSF and spinal regions and should be taken into consideration when designing intrathecal imaging studies. A potentially interesting relationship was observed between absorbed dose and administered volume, which merits further investigation.

Automated segmentation of MR imaging to determine normative central nervous system cerebrospinal fluid volumes in healthy volunteers.

An accurate non-invasive method to determine total body cerebrospinal fluid volume has a number of potential diagnostic and therapeutic applications. Herein we describe a technique for automated segmentation of total body MRI data to determine cranial and spinal CSF volume in 15 healthy adults. These in vivo estimates of CSF volume exceed the standard reported volume of 150mL in human adults and provide normative data for diagnosis of disease states such as hydrocephalus and therapy including pharmacologic dosimetry. No correlation was observed between patient height or weight and total body CSF volume.

Dynamic dual-isotope molecular imaging elucidates principles for optimizing intrathecal drug delivery.

The intrathecal (IT) dosing route offers a seemingly obvious solution for delivering drugs directly to the central nervous system. However, gaps in understanding drug molecule behavior within the anatomically and kinetically unique environment of the mammalian IT space have impeded the establishment of pharmacokinetic principles for optimizing regional drug exposure along the neuraxis. Here, we have utilized high-resolution single-photon emission tomography with X-ray computed tomography to study the behavior of multiple molecular imaging tracers following an IT bolus injection, with supporting histology, autoradiography, block-face tomography, and MRI. Using simultaneous dual-isotope imaging, we demonstrate that the regional CNS tissue exposure of molecules with varying chemical properties is affected by IT space anatomy, cerebrospinal fluid (CSF) dynamics, CSF clearance routes, and the location and volume of the injected bolus. These imaging approaches can be used across species to optimize the safety and efficacy of IT drug therapy for neurological disorders.

Regional, kinetic [(18)F]FDG PET imaging of a unilateral Parkinsonian animal model.

Positron emission tomography (PET) imaging with the glucose analog 2-deoxy-2-[(18)F]fluoro-D-glucose ([(18)F] FDG) has demonstrated clinical utility for the monitoring of brain glucose metabolism alteration in progressive neurodegenerative diseases. We examined dynamic [(18)F]FDG PET imaging and kinetic modeling of atlas-based regions to evaluate regional changes in the cerebral metabolic rate of glucose in the widely-used 6-hydroxydopamine (6-OHDA) rat model of Parkinson's disease. Following a bolus injection of 18.5 ± 1 MBq [(18)F]FDG and a 60-minute PET scan, image-derived input functions from the vena cava and left ventricle were used with three models, including Patlak graphical analysis, to estimate the influx constant and the metabolic rate in ten brain regions. We observed statistically significant changes in [(18)F]FDG uptake ipsilateral to the 6-OHDA injection in the basal ganglia, olfactory bulb, and amygdala regions; and these changes are of biological relevance to the disease. These experiments provide further validation for the use of [(18)F]FDG PET imaging in this model for drug discovery and development.

Regional particle size dependent deposition of inhaled aerosols in rats and mice.

CONTEXT: The current data analysis tools in nuclear medicine have not been used to evaluate intra organ regional deposition patterns of pharmaceutical aerosols in preclinical species. OBJECTIVE: This study evaluates aerosol deposition patterns as a function of particle size in rats and mice using novel image analysis techniques. MATERIALS AND METHOD: Mice and rats were exposed to radiolabeled polydisperse aerosols at 0.5, 1.0, 3.0, and 5.0 µm MMAD followed by SPECT/CT imaging for deposition analysis. Images were quantified for both macro deposition patterns and regional deposition analysis using the LRRI-developed Onion Model. RESULTS: The deposition fraction in both rats and mice was shown to increase as the particle size decreased, with greater lung deposition in rats at all particle sizes. The Onion Model indicated that the smaller particle sizes resulted in increased peripheral deposition. DISCUSSION: These data contrast the commonly used 10% deposition fraction for all aerosols between 1.0 and 5.0 µm and indicate that lung deposition fraction in this range does change with particle size. When compared to historical data, the 1.0, 3.0, and 5.0 µm particles result in similar lung deposition fractions; however, the 0.5 µm lung deposition fraction is markedly different. This is probably caused by the current aerosols that were polydisperse to reflect current pharmaceutical aerosols, while the historical data were generated with monodisperse aerosols. CONCLUSION: The deposition patterns of aerosols between 0.5 and 5.0 µm showed an increase in both overall and peripheral deposition as the particle size decreased. The Onion Model allows a more complex analysis of regional deposition in preclinical models.

Assessing antibody pharmacokinetics in mice with in vivo imaging.

Recent advances in small-animal molecular imaging instrumentation combined with well characterized antibody-labeling chemistry have enabled detailed in vivo measurements of antibody distribution in mouse models. This article reviews the strengths and limitations of in vivo antibody imaging methods with a focus on positron emission tomography and single-photon emission computed tomography and a brief discussion of the role of optical imaging in this application. A description of the basic principles behind the imaging techniques is provided along with a discussion of radiolabeling methods relevant to antibodies. Practical considerations of study design and execution are presented through a discussion of sensitivity and resolution tradeoffs for these techniques as defined by modality, signaling probe (isotope or fluorophore) selection, labeling method, and radiation dosimetry. Images and analysis results from a case study are presented with a discussion of output data content and relevant informatics gained with this approach to studying antibody pharmacokinetics.

Maximum-Likelihood Estimation With a Contracting-Grid Search Algorithm.

A fast search algorithm capable of operating in multi-dimensional spaces is introduced. As a sample application, we demonstrate its utility in the 2D and 3D maximum-likelihood position-estimation problem that arises in the processing of PMT signals to derive interaction locations in compact gamma cameras. We demonstrate that the algorithm can be parallelized in pipelines, and thereby efficiently implemented in specialized hardware, such as field-programmable gate arrays (FPGAs). A 2D implementation of the algorithm is achieved in Cell/BE processors, resulting in processing speeds above one million events per second, which is a 20× increase in speed over a conventional desktop machine. Graphics processing units (GPUs) are used for a 3D application of the algorithm, resulting in processing speeds of nearly 250,000 events per second which is a 250× increase in speed over a conventional desktop machine. These implementations indicate the viability of the algorithm for use in real-time imaging applications.

Adaptive SPECT.

Adaptive imaging systems alter their data-acquisition configuration or protocol in response to the image information received. An adaptive pinhole single-photon emission computed tomography (SPECT) system might acquire an initial scout image to obtain preliminary information about the radiotracer distribution and then adjust the configuration or sizes of the pinholes, the magnifications, or the projection angles in order to improve performance. This paper briefly describes two small-animal SPECT systems that allow this flexibility and then presents a framework for evaluating adaptive systems in general, and adaptive SPECT systems in particular. The evaluation is in terms of the performance of linear observers on detection or estimation tasks. Expressions are derived for the ideal linear (Hotelling) observer and the ideal linear (Wiener) estimator with adaptive imaging. Detailed expressions for the performance figures of merit are given, and possible adaptation rules are discussed.

Hardware assessment using the multi-module, multi-resolution system (M3R): a signal-detection study.

The multi-module, multi-resolution system (M3R) is used for hardware assessment in objective, task-based signal detection studies in projection data. A phantom capable of generating multiple realizations of a random textured background is introduced. Measured backgrounds from this phantom are used along with simulated lumpy and uniform backgrounds to investigate signal-to-noise ratio as a function of exposure time. Results are shown to agree with theoretical predictions, exhibiting a power-law like dependence previously seen for studies performed either in simulation or without an imaging system, and help validate the use of simulated lumpy backgrounds in observer studies. A second study looks at signal-detection performance, measured by AUC (area under the receiver operating characteristic curve), in lumpy backgrounds for 20 M3R aperture combinations as a function of lump size and signal size. Observer performance reveals an improvement in AUC for certain ranges of signal and lump combinations through the use of multiplexed, multiple-pinhole apertures, indicating a need for task-specific aperture optimization. The channelized Hotelling observer is used with Laguerre-Gauss channels for both observer studies. Methods for selection of number of channels and channel width are discussed.

The multi-module, multi-resolution system (M3R): a novel small-animal SPECT system.

We have designed and built an inexpensive, high-resolution, tomographic imaging system, dubbed the multi-module, multi-resolution system, or M3R. Slots machined into the system shielding allow for the interchange of pinhole plates, enabling the system to operate over a wide range of magnifications and with virtually any desired pinhole configuration. The flexibility of the system allows system optimization for specific imaging tasks and also allows for modifications necessary due to improved detectors, electronics, and knowledge of system construction (e.g., system sensitivity optimization). We provide an overview of M3R, focusing primarily on system design and construction, aperture construction, and calibration methods. Reconstruction algorithms will be described and reconstructed images presented.

Evaluation of hardware in a small-animal SPECT system using reconstructed images.

Evaluation of imaging hardware represents a vital component of system design. In small-animal SPECT imaging, this evaluation has become increasingly difficult with the emergence of multi-pinhole apertures and adaptive, or patient-specific, imaging. This paper will describe two methods for hardware evaluation using reconstructed images. The first method is a rapid technique incorporating a system-specific non-linear, three-dimensional point response. This point response is easily computed and offers qualitative insight into an aperture's resolution and artifact characteristics. The second method is an objective assessment of signal detection in lumpy backgrounds using the channelized Hotelling observer (CHO) with 3D Laguerre-Gauss and difference-of-Gaussian channels to calculate area under the receiver-operating characteristic curve (AUC). Previous work presented at this meeting described a unique, small-animal SPECT system (M(3)R) capable of operating under a myriad of hardware configurations and ideally suited for image quality studies. Measured system matrices were collected for several hardware configurations of M(3)R. The data used to implement these two methods was then generated by taking simulated objects through the measured system matrices. The results of these two methods comprise a combination of qualitative and quantitative analysis that is well-suited for hardware assessment.

Experimental task-based optimization of a four-camera variable-pinhole small-animal SPECT system.

We have previously utilized lumpy object models and simulated imaging systems in conjunction with the ideal observer to compute figures of merit for hardware optimization. In this paper, we describe the development of methods and phantoms necessary to validate or experimentally carry out these optimizations. Our study was conducted on a four-camera small-animal SPECT system that employs interchangeable pinhole plates to operate under a variety of pinhole configurations and magnifications (representing optimizable system parameters). We developed a small-animal phantom capable of producing random backgrounds for each image sequence. The task chosen for the study was the detection of a 2mm diameter sphere within the phantom-generated random background. A total of 138 projection images were used, half of which included the signal. As our observer, we employed the channelized Hotelling observer (CHO) with Laguerre-Gauss channels. The signal-to-noise (SNR) of this observer was used to compare different system configurations. Results indicate agreement between experimental and simulated data with higher detectability rates found for multiple-camera, multiple-pinhole, and high-magnification systems, although it was found that mixtures of magnifications often outperform systems employing a single magnification. This work will serve as a basis for future studies pertaining to system hardware optimization.