Positron emission tomography is a nuclear imaging technique that allows for assessment of physiologic processes and tissue metabolism after radionuclide administration. The most commonly used PET radiopharmaceutical is 18FDG, which is used to evaluate glucose metabolism.
As a glucose analog, 18FDG is taken into the cell via the same transport mechanisms as glucose. Once inside the cell, 18FDG is phosphorylated by hexokinase resulting in 18FDG-6-phosphate; because of the substitution of 18F in the C2 position of the deoxyglucose molecule, 18FDG-6-phosphate cannot undergo further glycolytic metabolism, thereby remaining trapped within the cell.1–3 Cells with high metabolic rates will have proportionally high concentrations of 18FDG, reflecting the rate of glucose utilization.1–3
Positron emission tomography with 18FDG is used in humans with intracranial disease to stage and aid in determining prognosis of primary brain tumors, differentiate between benign and malignant neoplasia, identify metastatic neoplasia, identify areas of inflammation, evaluate response to treatment, and evaluate patients with epilepsy.1–10 The development of 18FDG PET for companion animals is still in its infancy; however, there are case reports of 18FDG being used for assessment of intracranial disease, including imaging of primary brain tumors, meningoencephalitis, and epilepsy, in dogs.11–13,a
Autoimmune or immune-mediated encephalitis has only recently been recognized in humans because of its rarity and the difficulty of obtaining a definitive diagnosis.5 Humans with autoimmune or immune-mediated encephalitis can have severe neurologic signs including seizures, movement disorders, and psychosis, but despite these neurologic signs, findings of brain MRI frequently appear normal.5–7 The use of 18FDG PET has been helpful in identifying patients with this type of encephalitis in the face of apparently normal MRI findings, including patients with neuropsychiatric lupus erythematous, anti-N-methyl-d-aspartate receptor encephalitis, paraneoplastic limbic encephalitis, and voltage-gated potassium channel antibody–associated limbic encephalitis.5–7,14,15 The use of 18FDG PET in these patients continues to grow, and metabolic uptake patterns in certain anatomic regions of the brain for specific types of encephalitis are being identified.7 Not only have 18FDG-PET data facilitated diagnosis of these diseases, but also have provided clinicians with the ability to monitor response to immunotherapy and to assess disease progression.5,7
Dogs can also develop encephalitis that is likely immune mediated in origin, including GME, NME, and necrotizing leukoencephalitis.16 Similar to autoimmune or immune-mediated encephalitis in humans, definitive diagnosis is often difficult and ultimately requires surgical brain biopsy and histologic examination of collected tissue specimens. The prognosis for dogs with GME is generally more favorable than that for dogs with NME,16–19 but antemortem differentiation between these 2 disease forms requires histologic analysis of brain tissue. To date, no reliable MRI patterns or CSF characteristics have been identified to aid clinicians in differentiation between the granulomatous and necrotizing forms of encephalitis. Positron emission tomography with 18FDG offers an attractive, non-invasive alternative for diagnosis of immune-mediated encephalitis in dogs with no abnormal findings on brain MRI or CSF analysis. Currently, data regarding 18FDG metabolism in brains of healthy dogs are scarce and brain anatomic ROIs have been poorly defined,3,20,21 which makes interpretation of data obtained from dogs affected with immune-mediated encephalitis difficult.
Previous studies3,20,21 in dogs have characterized 18FDG uptake by calculating the SUV of the whole brain or broad ROIs from late static images obtained at a fixed time point. Standardized uptake value calculation is influenced by many factors, including size and placement of the ROI, imaging protocol and algorithm used for image reconstruction, body composition, blood glucose concentration, and change in uptake over time; thus, use of SUV measurement calculated at a single time point makes comparison among patients difficult.22 Kinetic analysis can provide greater insight into the behavior of 18FDG in specific regions and has the potential to provide more accurate characterizations of uptake of the compound in tissue and of glucose metabolism. Kinetic analysis can also assist in determining the time points at which static images that accurately assess glucose uptake and distribution within ROIs in the brain should be acquired. Previous research in juvenile dogs demonstrated little change in visceral organ SUVs between 1 and 2 hours after injection of 18FDG, although in that study,23 the brain was not included in the scanner field of view and, thus, could not be assessed dynamically. The purpose of the study reported here was to assess kinetic 18FDG uptake in the brain in anesthetized healthy adult dogs by use of PET and to determine whether 18FDG uptake differs among anatomic regions of the brain.
Materials and Methods
Five healthy sexually intact male Beagles weighing 8.6 to 10.9 kg and 3 to 5 years of age were used. Results of physical and neurologic examination performed 2 days prior to the imaging procedures were considered normal for all dogs. A CBC and serum biochemical analysis with electrolyte assessments performed 3 weeks prior to the imaging procedures revealed no abnormalities. The dogs were housed in a University of Tennessee College of Veterinary Medicine Animal and Care Committee–approved facility and used in accordance to a protocol approved by the Institutional Animal Care and Use Committee. On the day of the PET scan, the dogs were transported to and from the Pre-clinical Diagnostic Imaging Center at the University of Tennessee Medical Center via a temperature-controlled motor vehicle in accordance with University of Tennessee Radiation Safety guidelines.
For 12 hours prior to induction of anesthesia, food was withheld from the dogs, but they were allowed free access to water. Each dog was premedicated with butorphanol tartrateb (0.3 mg/kg, IV) and acepromazine maleatec (0.01 mg/kg, IV) approximately 15 minutes prior to IV catheter placement. Following catheter placement, propofold (3 mg/kg, IV) was used for induction of anesthesia, followed by isofluranee administered to effect for maintenance of anesthesia. Dogs received saline (0.9% NaCl) solution (4 mL/kg/h) during the scanning procedure. Blood glucose concentration was measured immediately prior to 18FDG administration.
All PET imaging was performed with a 64-slice, 4-ring PET-CT scanner.f Acquisition of PET data began several seconds prior to injection of 18FDG via the peripheral IV catheter (mean ± SD dose of 18FDG, 4.2 ± 1.1 mCi). The maximum and minimum dose of 18FDG received by any dog was 5.2 and 3.0 mCi, respectively. Data were acquired for 2 hours and dynamically framed into 46 frames. Only the first 38 frames of data (90-minute period) were found to contain the critical information for this analysis. All PET data were corrected for radioactive decay, attenuation, and scatter. The data were reconstructed with manufacturer-provided processing softwareg on the basis of ordered subset expectation maximization with point spread function correction that included 3 iterations and 24 subsets. The reconstructed image dimensions were 256 × 256 × 112, resulting in a reconstructed voxel size of 1.59 × 1.59 × 2.0 mm. For each dog, a non–contrast-enhanced CT scan was performed before PET acquisition for confirmation of the dogs’ positioning and for gross anatomic localization and attenuation correction.
All dogs recovered uneventfully from anesthesia and were transported back to the approved housing facility where they were quarantined until the external radiation exposure rate decreased to acceptable levels, as determined by the University of Tennessee Radiation Safety guidelines (< 0.2 mR/h at distance of 1 m from the animal), whereupon the dog was released from quarantine.
After image reconstruction, ROIs were drawn with postprocessing computer software.h By use of fused PET and CT images for more precise anatomic localization, 18FDG uptake was quantified by drawing ROIs over the following 15 areas of the brain: olfactory bulbs, gray matter of left and right frontal lobes, white matter of left and right frontal lobes, gray matter of left and right parietal lobes, white matter of left and right parietal lobes, left and right occipital lobes, brainstem, left and right cerebellar lobes, and midcerebellum at the level of the vermis. Subsequently, data were derived for the frontal and parietal lobes from the gray and white matter components of each. The ROIs were drawn with free-hand technique by a veterinary neurology resident (LMW) supervised by a board-certified veterinary radiologist (FM) with > 6 years of experience in PET imaging. The ROIs were drawn on dorsal images, with the exception of ROIs for the olfactory bulbs (for which transverse images were used) and brain-stem and cerebellum (for which sagittal images were used; Figure 1). The ROI placement was then cross-checked on the other image planes and adjusted as needed. Standard uptake values were calculated automatically from the acquisition software as (ROI activity × body weight)/injected dose, where ROI activity is measured in becquerels per milliliter, body weight is measured in g, and injected dose is measured in becquerels.
Representative dorsal plane PET image of the head of a healthy Beagle to illustrate placement of 15 ROIs over the following areas of the brain: olfactory bulbs, gray matter of left and right frontal lobes, white matter of left and right frontal lobes, gray matter of left and right parietal lobes, white matter of left and right parietal lobes, left and right occipital lobes, brainstem, left and right cerebellar lobes, and midcerebellum at the level of the vermis. Regions of interest were drawn on dorsal images, with the exception of the olfactory bulbs (for which transverse images were used) and brainstem and cerebellum (for which sagittal images were used); ROI placement was then crosschecked against CT image planes and adjusted as needed.
Citation: American Journal of Veterinary Research 75, 6; 10.2460/ajvr.75.6.588
Image data analysis—For the regions that had measurements obtained from the left and right hemisphere components, mean SUV at each time point for each region was calculated. Left and right regions were combined in this manner for the following regions: cerebellum (including data for the midcerebellum at the level of the vermis), parietal lobes, frontal lobes, occipital lobes, and white and gray matter. Time activity curves were generated with postprocessing computer softwareh and exported for analysis in spreadsheet software.i Data from each sample were collated into collective descriptive statistics to obtain population demographics (maximum SUV, time to maximum uptake, time rate of change of uptake, and maximum activity concentration) for the kinetics of each segmented region of the brain. Analyses of time to maximum measured activity concentration, defined as the time required to reach the maximum measured activity concentration during the acquisition time, and clearance rates were performed for the kinetic assessment of these data.
All voxel values used to generate TACs were saved (as Bq/mL and SUVs) with subsequent analysis after SUVs were normalized to cerebellar activity. This analysis helps to assess between-animal variations that can occur with regard to injected dose, dog weight, and brain uptake. For basic compartmental analysis to determine clearance rate estimates, the activity values used were in units of becquerels per milliliter. For each key region, data were also analyzed (with commercially available statistical softwarej) among dogs to evaluate the TACs and to determine Pearson correlation coefficients with linear regression methods along with 95% confidence intervals. Significance was set at a value of P ≤ 0.05.
First-order estimations of uptake rate and clearance rate were performed by compartmental modeling with 2 compartments and the linear least squares fitting method.24 Modeling was performed with modeling softwarek with variables obtained for estimated rate of uptake (K1) and estimated rate of clearance (k2). For 18FDG, the physical half-life is 110 minutes and the physical decay constant, Λp, is 0.0063 minutes−1. The half-life of the compound within the body is affected by biological processes, with a distinct biological decay constant, Λb. The sum of the physical and biological decay constants is the total effective decay constant. Logarithmic fitting was performed to the end time points of the TAC data to approximate elimination rate constants. The effective half-life was then calculated as ln 2/(Λp + Λb).
For each left-right pair of ROIs drawn, an AI was calculated to estimate any lateralization of changes in the regions:
Asymmetry indices for the ROI pairs drawn in the left and right hemispheres were then compared. Based on previously published data,25 only AIs > 9% or < −9% were considered significant.
Results
In all 5 dogs, mean ± SD blood glucose concentration (97.6 ± 6.2 mg/dL) was within reference range (70 to 138 mg/dL). Time-activity curve analysis clearly identified different patterns of uptake for certain regions and groups of regions within the 15 areas segmented for assessment in the study. For all dogs, 4 distinct groupings were visually seen when TACs were plotted on the same graph (Figure 2): olfactory regions, gray matter regions, white matter regions, and all other regions (brainstem, cerebellum [including the midcerebellum at the level of the vermis], and occipital and frontal lobes).
Time-activity curves (displayed as means of PET data acquired over a 90-minute period from the 15 ROIs described in Figure 1) for uptake of 18FDG following IV administration (dose range, 3.0 to 5.2 mCi) in 5 anesthetized healthy dogs. Four distinct groupings are apparent when all TACs are plotted on the same graph: olfactory regions, gray matter regions, white matter regions, and all other regions (includes brainstem, cerebellum [including the midcerebellum at the level of the vermis], occipital, and frontal regions).
Citation: American Journal of Veterinary Research 75, 6; 10.2460/ajvr.75.6.588
The olfactory regions had the highest degree of uptake, with a mean time to peak uptake of 82.7 minutes; this indicates that uptake in this region continued to increase through the entire 90-minute acquisition period (ie, the first 90 minutes of a 2-hour acquisition period). White and gray matter regions differed significantly (P < 0.05) in time to peak uptake, with a mean of 37.8 and 67.7 minutes, respectively. Occipital, brainstem, and all cerebellar regions were all approximately equivalent, with time to peak uptake between 42.8 and 47.8 minutes. Certain regions were also found to have identical times to peak uptake: 47.8 minutes for the brainstem and occipital regions and 62.7 minutes for the parietal and frontal lobe regions. The maximum activity concentration, time to peak uptake, and the rate of change in maximum SUV for each region were summarized (Table 1).
Summary of mean uptake variables derived for 15 ROIs in the brains of 5 healthy Beagles that were administered 18FDG IV (dose range, 3.0 to 5.2 mCi) and underwent PET.
| Region | Time to peak uptake (min) | Rate of change in SUV (ΔSUV/Δmin) | Maximum activity concentration (kBq/mL) | Maximum SUV |
|---|---|---|---|---|
| Olfactory | 82.7 | 0.094 | 135.6 | 7.87 |
| Cerebellum | 42.8 | 0.108 | 84.4 | 4.86 |
| Parietal | 62.7 | 0.074 | 83.8 | 4.82 |
| Frontal | 62.7 | 0.068 | 77.1 | 4.45 |
| Occipital | 47.8 | 0.098 | 83.3 | 4.88 |
| Brainstem | 47.8 | 0.092 | 86.2 | 4.89 |
| White matter | 37.8 | 0.087 | 60.7 | 3.44 |
| Gray matter | 67.7 | 0.084 | 100.9 | 5.87 |
| Overall mean | 56.5 | 0.088 | 86.6 | 5.14 |
Data were collected for each ROI over a 2-hour period; for analysis, data collected during the first 90 minutes were used. Placement of ROIs was as follows: olfactory bulbs, gray matter of left and right frontal lobes, white matter of left and right frontal lobes, gray matter of left and right parietal lobes, white matter of left and right parietal lobes, left and right occipital lobes, brainstem, left and right cerebellar lobes, and midcerebellum at the level of the vermis. Subsequently, data were derived for the frontal and parietal lobes from the gray and white matter components of each. Regions of interest were drawn on dorsal images, with the exception of the olfactory bulbs (for which transverse images were used) and brainstem and cerebellum (for which sagittal images were used); ROI placement was then cross-checked on the other image planes and adjusted as needed. The arithmetic mean of the data for the left and right hemisphere components of a region was calculated for purposes of analysis.
Linear least squares fitting for a 2-compartment model revealed a mean uptake rate for all regions of 8.6 min−1 (Table 2). The mean clearance rate for all regions was 1.13 minutes−1. Among the higher uptake rates, confirmed visually and quantitatively, the olfactory region had the maximum rate of 9.82 mL/min, whereas the region with the lowest rate of uptake was the white matter (6.44 mL/min). The cerebellar regions showed the fastest clearance (1.34 minutes−1), and the olfactory region had the slowest clearance (0.72 minutes−1). The K1:k2 ratios confirm quantitatively the delineation of the 4 region groupings identified in visual TAC analysis. The olfactory region had the highest K1:k2 ratio (13.71), followed by the gray matter (9.52). The regions grouped as other had a K1:k2 ratio of approximately 7, and white matter had the lowest ratio (5.17).
Summary of mean 18FDG uptake (K1) and clearance (k2) rates for each ROI of the brains of the 5 healthy Beagles in Table 1.
| Region | K1 (min−1) | k2 (min−1) | K1:k2 ratio |
|---|---|---|---|
| Olfactory | 9.82 | 0.72 | 13.71 |
| Cerebellum | 9.41 | 1.34 | 7.03 |
| Parietal | 8.12 | 1.08 | 7.55 |
| Frontal | 7.70 | 1.10 | 6.99 |
| Occipital | 9.32 | 1.32 | 7.07 |
| Brainstem | 8.67 | 1.22 | 7.10 |
| White matter | 6.44 | 1.25 | 5.17 |
| Gray matter | 9.41 | 0.99 | 9.52 |
| Overall mean | 8.61 | 1.13 | 8.02 |
Logarithmic fitting of the TAC data to the end time points yielded approximations of elimination rate constants with a mean for all regions of 0.00817 minutes−1. The effective half-life was then calculated to be 84.8 minutes. The greatest effective half-life in the olfactory region (102.2 minutes) and the lowest half-life in the occipital region (74.1 minutes). Biological decay constants and effective half-lives were summarized (Table 3).
Summary of mean biological decay constants and effective half-lives of 18FDG in the brains of the 5 healthy Beagles in Table 1.
| Region | Biological decay constant (min−1) | Effective half-life (min) |
|---|---|---|
| Olfactory | 4.8 × 10−4 | 102.2 |
| Cerebellum | 2.3 × 10−3 | 81.0 |
| Parietal | 1.9 × 10−3 | 84.9 |
| Frontal | 1.6 × 10−3 | 88.0 |
| Occipital | 3 × 10−3 | 74.1 |
| Brainstem | 2.7 × 10−3 | 77.4 |
| White matter | 2.2 × 10−3 | 81.2 |
| Gray matter | 1.9 × 10−3 | 89.7 |
Examination of TACs drawn in the left and right cerebral hemispheres revealed minimal difference in time to maximum uptake estimations (Table 4). The mean difference in time to maximum uptake estimations between left and right regions of the brain was only 5.81 minutes. Linear regression analysis of the mean values for the left and right cerebral hemisphere ROIs yielded a mean correlation coefficient (R2) of 0.997, with a 95% confidence interval of 0.997 to 1.035. Analysis of asymmetry revealed no significant (P > 0.05) variation between cerebral left and right hemispheres, with all AIs < 5.4%. Results of the analysis of negative and positive relationships indicated that there was a slight tendency for increased uptake in the right hemisphere, compared with uptake in the left hemisphere. Gray matter regions had had the greatest degree of left lateralization (−5.32%); white matter had the greatest degree of right lateralization (3.82%).
Summary of differences and AIs for 18FDG between the left and left hemispheres ROIs in the brains of the 5 healthy Beagles in Table 1.
| Region | R2 | P value | Time to maximum activity concentration (min) | Mean difference between left and right hemisphere activity concentration measurements (%) | AI (%) |
|---|---|---|---|---|---|
| Cerebellum | 0.995 | 0.000 | 4.98 | 8.52 | −0.085 |
| Parietal | 0.999 | 0.000 | 0.00 | 2.96 | 0.347 |
| Frontal | 0.997 | 0.001 | 4.98 | 2.59 | −4.376 |
| Occipital | 0.995 | 0.000 | 14.95 | 5.23 | −0.365 |
| White matter | 0.997 | 0.000 | 4.98 | 0.12 | 3.816 |
| Gray matter | 0.998 | 0.000 | 4.98 | 4.95 | −5.321 |
| Overall mean | 0.997 | 0.000 | 5.81 | 4.02 | NA |
NA = Not applicable.
Discussion
In the present study involving anesthetized healthy adult dogs, the highest maximum SUVs for 18FDG (determined by PET) were observed in the olfactory bulbs and cerebral gray matter. The high uptake in the olfactory bulbs was expected because dogs are a macrosmatic species with a highly developed sense of smell.26 Also, gray matter consists of neurons and glial cells, which are the metabolically active components of the brain. Results of a previous study20 assessing 18FDG uptake in the brain 20 minutes after administration to 5 healthy Beagles indicated that the highest uptakes were within the frontal and occipital lobes. This discrepancy in findings is likely attributable to the other researchers’ inclusion of the olfactory bulb within the frontal lobe ROIs. In the present study, cerebral white matter had the lowest 18FDG uptake, which is explained by the lack of neurons and the small amount of metabolically active glial cells in healthy white matter.
With regard to time to achieve peak uptake, the most metabolically active area of the brain (the olfactory region) required a mean interval of 83 minutes, and the remaining brain tissues required a mean interval of 68 minutes or less. If neurologic signs or MRI findings do not indicate involvement of the olfactory lobe, a scan duration of approximately 1 hour should suffice to obtain pertinent 18FDG uptake data.
Although CT and MRI are classically regarded as the primary methods of neuroimaging, those techniques are somewhat limited by their inability to detect functional alterations prior to structural changes. Positron emission tomography with administration of 18FDG can fill the diagnostic void left by traditional neuroimaging methods. Changes in glucose uptake in tissues, represented by 18FDG hypermetabolism or hypometabolism, can be indicative of neurologic disease at a functional level. Hypermetabolism can occur with enhanced activity of inflammatory cells or reduced inhibition of neuronal cells.6 Specifically, activated immune cells (eg, T lymphocytes, macrophages, and neutrophils) can have enhanced glycolytic rates.8 Areas of hypometabolism can indicate tissue atrophy, infarction, or decreased tissue perfusion and decreased density of neuronal cell bodies.6
Dogs with GME may have no abnormal brain MRI findings and CSF variables may be within reference limits; thus, collection of brain biopsy specimens for examination may be the only potential means for definitive diagnosis.19 Minimally invasive brain biopsy procedures do not always yield a diagnostically useful sample; in a recent report,27 45 samples of forebrain lesions in 17 dogs were examined and 18% had no diagnostic value. Furthermore, a morbidity rate of 29% was associated with those minimally invasive brain biopsy procedures.27 Imaging of neuroinflammation by means of 18FDG PET could offer an alternative to tissue sample collection and would be especially useful in cases where a brain biopsy is declined by the owner or when the dog cannot undergo an invasive procedure involving general anesthesia.
The potential for identification of disease-specific 18FDG-PET uptake patterns in dogs, which would perhaps enable discrimination of the granulomatous and necrotizing forms of encephalitis, is also appealing and would provide veterinarians with the ability to offer more accurate prognostic information to dog owners. Two dogs with histologically confirmed NME described in a case report13 had areas of hypometabolism detected via 18FDG-PET imaging, and those areas corresponded to zones of malacia or brain atrophy. Malacia is not a reported feature of GME, so this could provide support for a noninvasive means for antemortem diagnosis of NME.
In the brains of the healthy dogs used in the present study, there was a significantly higher 18FDG uptake in the gray matter, compared with uptake in the white matter. Although brain lesion distribution is not always pathognomonic for differentiation of NME and GME, there are noticeable trends in lesion distribution. Dogs with GME generally have more intensive involvement of the white matter of the cerebrum, cerebellum, and brainstem, whereas dogs with NME generally have involvement of both the gray and white matter of the cerebrum.28,29
To our knowledge, this is the first study of 18FDG kinetics in a subset of anesthetized clinically normal dogs; these data provide the baseline for further investigation of 18FDG-PET uptake in dogs with inflammatory brain disease. Such future studies could clarify the role of 18FDG PET in the diagnosis of dogs with neuroinflammation, especially patients with apparently normal findings on MRI or CSF analysis and those for which a histopathologic diagnosis cannot be obtained. In addition, prospective clinical trials are needed to determine the role of 18FDG PET in differentiating between GME and NME and to evaluate response to treatment in dogs with encephalitis.
ABBREVIATIONS
| 18FDG | 2-([18F]fluoro)-2-deoxy-d-glucose |
| AI | Asymmetry index |
| GME | Granulomatous meningoencephalomyelitis |
| NME | Necrotizing meningoencephalitis |
| PET | Positron emission tomography |
| ROI | Region of interest |
| SUV | Standardized uptake value |
| TAC | Time-activity curve |
Viitmaa R, Haaparanta M, Cizinauskas S, et al. FDG-PET in normal and epileptic Finnish Spitz Dogs (abstr), in Proceedings. 12th Annu Conf Eur Assoc Vet Diagn Imag 2005;44.
Torbugesic, Fort Dodge Laboratories, Fort Dodge, Iowa.
Acepromazine maleate, Vedco Inc, St Joseph, Mo.
Propoflo, Abbott Laboratories, Chicago, Ill.
Isoflo, Abbott Laboratories, Chicago, Ill.
Biograph mCT, Siemens Molecular Imaging USA Inc, Malvern, Pa.
HD•PET, Siemens Molecular Imaging USA Inc, Malvern, Pa.
General Analysis Tool, Inveon Research Workplace, version 4.0, Siemens Medical Solutions USA Inc, Malvern, Pa.
Excel 2013, Microsoft Corp, Redmond, Wash.
SPSS Statistics, version 21, IBM Corp, Armonk, NY.
PMOD, version 3.2, PMOD Technologies LTD, Zurich, Switzerland.
References
1. Blodgett TM, Meltzer CC, Townsend DW. PET/CT: form and function. Radiology 2007; 242: 360–385.
2. Cohade C, Wahl RL. Applications of positron emission tomography/computed tomography image fusion in clinical positron emission tomography—clinical use, interpretation methods, diagnostic improvements. Semin Nucl Med 2003; 33: 228–237.
3. Lee MS, Lee AR, Jung MA, et al. Characterization of physiologic 18F-FDG uptake with PET-CT in dogs. Vet Radiol Ultrasound 2010; 51: 670–673.
4. Gok B, Jallo G, Hayeri R, et al. The evaluation of FDG-PET imaging for epileptogenic focus localization in patients with MRI positive and MRI negative temporal lobe epilepsy. Neuroradiology 2013; 55: 541–550.
5. Sekigawa M, Okumura A, Niijima S, et al. Autoimmune focal encephalitis shows marked hypermetabolism on positron emission tomography. J Pediatr 2010; 156: 158–160.
6. Lee SW, Park MC, Lee SK, et al. The efficacy of brain 18F-flurodeoxyglucose positron emission tomography in neuropsychiatric lupus patients with normal brain magnetic resonance imaging findings. Lupus 2012; 21: 1531–1537.
7. Leypoldt F, Buchert R, Kleiter I, et al. Flurodeoxyglucose positron emission tomography in anti-N-methyl-d-aspartate receptor encephalitis: distinct pattern of disease. J Neurol Neurosurg Psychiatry 2012; 83: 681–686.
8. Wunder A, Klohs J, Dirnagl U. Non-invasive visualization of CNS inflammation with nuclear and optical imaging. Neuroscience 2009; 158: 1161–1173.
9. Santra A, Kumar R, Sharma P, et al. F-18 FDG PET-CT for predicting survival in patients with recurrent glioma: a prospective study. Neuroradiology 2011; 53: 1017–1024.
10. Das K, Mittal BR, Vasistha RK, et al. Role of 18F-flurodeoxyglucose positron emission tomography scan in differentiating enhancing brain tumors. Indian J Nucl Med 2011; 26: 171–176.
11. Kang BT, Park C, Yoo JH, et al. 18F-fluorodeoxyglucose positron emission tomography and magnetic resonance imaging findings of primary intracranial histiocytic sarcoma in a dog. J Vet Med Sci 2009; 71: 1397–1401.
12. Kang BT, Kim SG, Lim CY, et al. Correlation between fluorodeoxyglucose positron emission tomography and magnetic resonance imaging findings of non-suppurative meningoencephalitis in 5 dogs. Can Vet J 2010; 51: 986–992.
13. Eom KD, Lim CY, Gu SH, et al. Positron emission tomography features of canine necrotizing meningoencephalitis. Vet Radiol Ultrasound 2008; 49: 595–599.
14. Chatzikonstantinou A, Szabo K, Ottomeyer C, et al. Successive affection of bilateral temporomesial structures in a case of non-paraneoplastic limbic encephalitis demonstrated by serial MRI and FDG-PET. J Neurol 2009; 256: 1753–1755.
15. Scheid R, Lincke T, Voltz R, et al. Serial 18F-fluoro-2-deoxy-d-glucose positron emission tomography and magnetic resonance imaging of paraneoplastic limbic encephalitis. Arch Neurol 2004; 61: 1785–1789.
16. Talarico LR, Schatzberg SJ. Idiopathic granulomatous and necrotising inflammatory disorders of the canine central nervous system: a review and future perspectives. J Small Anim Pract 2010; 51: 138–149.
17. Jurney CH, Van Winkle TJ, Shofer FS, et al. Necrotizing encephalitis: a retrospective study of 36 cases. J Vet Intern Med 2007; 21: 641.
18. Muñana KR, Luttgen PJ. Prognostic factors for dogs with GME: 42 cases (1982–1996). J Am Vet Med Assoc 1998; 212: 1902–1906.
19. Granger N, Smith PM, Jeffery ND. Clinical findings and treatment of non-infectious meningoencephalomyelitis in dogs: a systematic review of 457 published cases. Vet J 2010; 184: 290–297.
20. Lee MS, Ko J, Lee AR, et al. Effects of anesthetic protocol on normal canine brain uptake of 18F-FDG assessed by PET/CT. Vet Radiol Ultrasound 2010; 51: 130–135.
21. Kang BT, Son YD, Lee SR, et al. FDG uptake of normal canine brain assessed by high-resolution research tomography—positron emission tomography and 7 T-magnetic resonance imaging. J Vet Med Sci 2012; 74: 1261–1267.
22. Hapdey S, Buvat I, Carson J, et al. Searching for alternatives to full kinetic analysis in 18F-FDG PET: an extension of the simplified kinetic analysis method [Erratum published in J Nucl Med 2011; 52:838]. J Nucl Med 2011; 52: 634–641.
23. LeBlanc AK, Jakoby B, Townsend DW, et al. Thoracic and abdominal organ uptake of 2-deoxy-2-[18F]fluro-d-glucose (18FDG) with positron emission tomography in the normal dog. Vet Radiol Ultrasound 2008; 49: 182–188.
24. Cai W, Feng D, Fulton R, et al. Generalized linear least squares algorithms for modeling glucose metabolism in the human brain with corrections for vascular effects. Comput Methods Programs Biomed 2002; 68: 1–14.
25. Savic I, Lindstrom P. PET and MRI show differences in cerebral asymmetry and functional connectivity between homo- and heterosexual subjects. Proc Natl Acad Sci U S A 2008; 105: 9403–9408.
26. de Lahunta A, Glass E. Visceral afferent systems. In: Veterinary neuroanatomy and clinical neurology. 3rd ed. St Louis: Saunders Elsevier, 2009; 444–445.
27. Flegel T, Oevermann A, Oechtering G, et al. Diagnostic yield and adverse effects of MRI-guided free-hand biopsies through a mini-burr hole in dogs with encephalitis. J Vet Intern Med 2012; 26: 969–976.
28. Flegel T, Henke D, Boettcher I, et al. Magnetic resonance imaging findings in histologically confirmed Pug dog encephalitis. Vet Radiol Ultrasound 2008; 49: 419–424.
29. Suzuki M, Uchida K, Morozumi M, et al. A comparative pathological study on canine necrotizing meningoencephalitis and granulomatous meningoencephalomyelitis. J Vet Med Sci 2003; 65: 1233–1239.
