Voxel-Based Analysis of [18F]-FDG Brain PET in Rats Using Data-Driven Normalization

Using the coefficient of determination to identify injury regions after stroke in pre-clinical FDG-PET images

BACKGROUND

In the analysis of brain fluorodeoxyglucose positron emission tomography (FDG-PET) images, intensity normalization is a necessary step to reduce inter-subject variability. However, the choice of the most appropriate normalization method in stroke studies remains unclear, as demonstrated by inconsistent findings in the literature.

MATERIALS AND METHODS

Here, we propose a regression- and single-subject-based model for analyzing FDG-PET images without intensity normalization. Two independent data sets were collected before and after middle cerebral artery occlusion (MCAO), with one comprising 120 rats and the other 96 rats. After data preprocessing, voxel intensities in the same region and hemisphere were paired before and after the MCAO scan. A linear regression model was applied to the paired data, and the coefficient of determination R2 was calculated to measure the linearity. The R2 values between the ipsilateral and contralateral hemispheres were compared, and significant regions were defined as those with reduced linearity. Our method was compared with voxel-wise analysis under different intensity normalization methods and validated using the triphenyl tetrazolium chloride (TTC) staining data.

RESULTS

The significant regions identified by the proposed method showed a large degree of overlap with the infarcted regions identified by TTC data, as measured by the Dice similarity coefficient (DSC). The average DSC of the proposed method was 59.7%, whereas the DSCs of the existing approaches ranged from 41.4%∼51.3%. Additional validation using receiver operating characteristic (ROC) demonstrated that the area under the curve (AUC) of the average ROC curves reached 0.84 using the proposed method, whereas existing methods achieved AUCs ranging from 0.77∼0.79. The identified regions were consistent across the two independent data sets, and some findings were corroborated by other publications.

CONCLUSIONS

The proposed model presents a novel quantitative approach for identifying injury regions post-stroke using FDG-PET images. The calculation does not require intensity normalization and can be applied to individual subjects. The method yields more sensitive results compared to existing identification methods.

Quantitative brain [18F]FDG PET beyond normal blood glucose levels

Introduction SUV measurements from static brain [18F]FDG PET acquisitions are a commonly used tool in preclinical research, providing a simple alternative for kinetic modelling, which requires complex and time-consuming dynamic acquisitions. However, SUV can be severely affected by the animal handling and preconditioning protocols, primarily by those that may induce changes in blood glucose levels (BGL). Here, we aimed at developing and investigating the feasibility of SUV-based approaches for a wide range of BGL far beyond normal values, and consequently, to develop and validate a new model to generate standardized and reproducible SUV measurements for any BGL. Material and methods We performed dynamic and static brain [18F]FDG PET acquisitions in 52 male Sprague-Dawley rats sorted into control (n = 10), non-fasting (n = 14), insulin-induced hypoglycemia (n = 12) and glucagon-induced hyperglycemia (n = 16) groups. Brain [18F]FDG PET images were cropped, aligned and co-registered to a standard template to calculate whole-brain and regional SUV. Cerebral Metabolic Rate of Glucose (CMRglc) was also estimated from 2-Tissue Compartment Model (2TCM) and Patlak plot for validation purposes. Results Our results showed that BGL=100±6 mg/dL can be considered a reproducible reference value for normoglycemia. Furthermore, we successfully established a 2nd-degree polynomial model (C1=0.66E-4, C2=-0.0408 and C3=7.298) relying exclusively on BGL measures at pre-[18F]FDG injection time, that characterizes more precisely the relationship between SUV and BGL for a wide range of BGL values (from 10 to 338 mg/dL). We confirmed the ability of this model to generate corrected SUV estimations that are highly correlated to CMRglc estimations (R2= 0.54 2TCM CMRgluc and R2= 0.49 Patlak CMRgluc). Besides, slight regional differences in SUV were found in animals from extreme BGL groups, showing that [18F]FDG uptake is mostly directed toward central regions of the brain when BGLs are significantly decreased. Conclusion Our study successfully established a non-linear model that relies exclusively on pre-scan BGL measurements to characterize the relationship between [18F]FDG SUV and BGL. The extensive validation confirmed its ability to generate SUV-based surrogates of CMRglu along a wide range of BGL and it holds the potential to be adopted as a standard protocol by the preclinical neuroimaging community using brain [18F]FDG PET imaging.

Assessment of a one-week ketogenic diet on brain glycolytic metabolism and on the status epilepticus stage of a lithium-pilocarpine rat model

The ketogenic diet (KD) has been shown to be effective in refractory epilepsy after long-term administration. However, its interference with short-term brain metabolism and its involvement in the early process leading to epilepsy remain poorly understood. This study aimed to assess the effect of a short-term ketogenic diet on cerebral glucose metabolic changes, before and after status epilepticus (SE) in rats, by using [18F]-FDG PET. Thirty-nine rats were subjected to a one-week KD (KD-rats, n = 24) or to a standard diet (SD-rats, n = 15) before the induction of a status epilepticus (SE) by lithium-pilocarpine administrations. Brain [18F]-FDG PET scans were performed before and 4 h after this induction. Morphological MRIs were acquired and used to spatially normalize the PET images which were then analyzed voxel-wisely using a statistical parametric-based method. Twenty-six rats were analyzed (KD-rats, n = 15; SD-rats, n = 11). The 7 days of the KD were associated with significant increases in the plasma β-hydroxybutyrate level, but with an unchanged glycemia. The PET images, recorded after the KD and before SE induction, showed an increased metabolism within sites involved in the appetitive behaviors: hypothalamic areas and periaqueductal gray, whereas no area of decreased metabolism was observed. At the 4th hour following the SE induction, large metabolism increases were observed in the KD- and SD-rats in areas known to be involved in the epileptogenesis process late-i.e., the hippocampus, parahippocampic, thalamic and hypothalamic areas, the periaqueductal gray, and the limbic structures (and in the motor cortex for the KD-rats only). However, no statistically significant difference was observed when comparing SD and KD groups at the 4th hour following the SE induction. A one-week ketogenic diet does not prevent the status epilepticus (SE) and associated metabolic brain abnormalities in the lithium-pilocarpine rat model. Further explorations are needed to determine whether a significant prevention could be achieved by more prolonged ketogenic diets and by testing this diet in less severe experimental models, and moreover, to analyze the diet effects on the later and chronic stages leading to epileptogenesis.

Do Tau Deposition and Glucose Metabolism Dissociate in Alzheimer's Disease Trajectory?

Background

Tau aggregation demonstrates close associations with hypometabolism in Alzheimer's disease (AD), although differing pathophysiological processes may underlie their development.

Objective

To establish whether tau deposition and glucose metabolism have different trajectories in AD progression and evaluate the utility of global measures of these pathological hallmarks in predicting cognitive deficits.

Methods

279 participants with amyloid-β (Aβ) status, and T1-weighted MRI scans, were selected from the Alzheimer's Disease Neuroimaging Initiative (http://adni.loni.usc.edu). We created the standard uptake value ratio images using Statistical Parametric Mapping 12 for [18F]AV1451-PET (tau) and [18F]FDG-PET (glucose metabolism) scans. Voxel-wise group and single-subject level SPM analysis evaluated the relationship between global [18F]FDG-PET and [18F]AV1451-PET depending on the Aβ status. Linear models assessed whether tau deposition or glucose metabolism better predicted clinical progression.

Results

There was a dissociation between global cerebral glucose hypometabolism and global tau load in amyloid-positive AD and amyloid-negative mild cognitive impairment (MCI) (p > 0.05). Global hypometabolism was only associated with global cortical tau in amyloid-positive MCI. Voxel-level single subject tau load better predicted neuropsychological performance, Alzheimer's disease assessment scale-cognitive (ADAS-Cog) 13 score, and one-year change compared with regional and global hypometabolism.

Conclusions

A dissociation between tau pathology and glucose metabolism at a global level in AD could imply that other pathological processes influence glucose metabolism. Furthermore, as tau is a better predictor of clinical progression, these processes may have independent trajectories and require independent consideration in the context of therapeutic interventions.

Evaluation the in vivo behaviors of PM2.5 in rats using noninvasive PET imaging with mimic particles

Inhaled PM2.5 particles is harmful to human health. However, real-time tracking of PM2.5 particles and dynamic evaluation of the pharmacokinetic behaviors in vivo are still challenging. Here, PET imaging is utilized to noninvasively monitor the in vivo behavior of PM2.5 particles in rats. To mimic aerosol PM2.5 particles suspended in ambient air, 89Zr-labeled melanin nanoparticles (89Zr-MNP) are nebulized into microscopic liquid particles with a mean size of 2.5 μm. Then, the 89Zr-labeled PM2.5 mimic particles (89Zr-PM2.5) are administrated into rats via inhalation. PET imaging showed that 89Zr-PM2.5 mainly accumulated in the lungs for up to 384 h after administration. Besides, we also observe that a small amount of 89Zr-PM2.5 can penetrate the brain through the inhalation. Further PET imaging showed that enhanced uptakes of 18F-FDG and 18F-DPA-714 were found in the brain of rats upon PM2.5 mimic particle exposure, which revealed that pulmonary exposure to PM2.5 could cause potential damages to the brain. Note that abnormal glucose metabolism was reversed, but the neuroinflammation was permanent and could not be alleviated after ceasing PM2.5 exposure. Our results demonstrate that PET is a sensitive and feasible tool for evaluating the in vivo behaviors of PM2.5.

Initial Experience on Hyperpolarized [1-13C]Pyruvate MRI Multicenter Reproducibility-Are Multicenter Trials Feasible?

BACKGROUND

Magnetic resonance imaging (MRI) with hyperpolarized [1-13C]pyruvate allows real-time and pathway specific clinical detection of otherwise unimageable in vivo metabolism. However, the comparability between sites and protocols is unknown. Here, we provide initial experiences on the agreement of hyperpolarized MRI between sites and protocols by repeated imaging of same healthy volunteers in Europe and the US.

METHODS

Three healthy volunteers traveled for repeated multicenter brain MRI exams with hyperpolarized [1-13C]pyruvate within one year. First, multisite agreement was assessed with the same echo-planar imaging protocol at both sites. Then, this was compared to a variable resolution echo-planar imaging protocol. In total, 12 examinations were performed. Common metrics of 13C-pyruvate to 13C-lactate conversion were calculated, including the kPL, a model-based kinetic rate constant, and its model-free equivalents. Repeatability was evaluated with intraclass correlation coefficients (ICC) for absolute agreement computed using two-way random effects models.

RESULTS

The mean kPL across all examinations in the multisite comparison was 0.024 ± 0.0016 s-1. The ICC of the kPL was 0.83 (p = 0.14) between sites and 0.7 (p = 0.09) between examinations of the same volunteer at any of the two sites. For the model-free metrics, the lactate Z-score had similar site-to-site ICC, while it was considerably lower for the lactate-to-pyruvate ratio.

CONCLUSIONS

Estimation of metabolic conversion from hyperpolarized [1-13C]pyruvate to lactate using model-based metrics such as kPL suggests close agreement between sites and examinations in volunteers. Our initial results support harmonization of protocols, support multicenter studies, and inform their design.