Evaluation of ¹⁸F-fluorodeoxyglucose uptake of beagle dogs for different durations of isoflurane anesthesia

Abstract

¹⁸F-fluorodeoxyglucose (FDG) positron emission tomography (PET) is used for tumor evaluation. In veterinary medicine, anesthesia is an essential tool during the PET scanning process. However, the changes in FDG uptake in dogs that have undergone anesthesia for a longer duration have not been studied. This study aimed to analyze the influence of isoflurane anesthesia on FDG uptake in dogs undergoing PET. A crossover design was implemented by exposing 3 groups of 6 dogs to different durations of anesthesia (60, 90, and 150 minutes). Inhalation anesthesia was maintained throughout the scanning process (30 minutes) and FDG was injected 1 hour before the start of the PET scan. The standard uptake value of FDG was obtained for the 7 gross structures (whole brain, lung, salivary gland, liver, spleen, mediastinal blood pool, and kidney cortex) as well as for the 7 intracranial structures (frontal, parietal, temporal and occipital lobes, cerebellum, brain stem, and caudal colliculus). The whole brain and intracranial structures showed significantly lower FDG uptake in dogs with a longer duration of anesthesia, whereas other gross structures did not. Our results suggest that the duration of anesthesia should be considered when evaluating the uptake of FDG by the brain.

Introduction

¹⁸F-fluorodeoxyglucose (FDG) is a radiopharmaceutical material that is widely used for the evaluation of tumor metabolism using positron emission tomography (PET) and computed tomography (CT). FDG-PET is commonly used to diagnose, stage, restage, and monitor neoplastic disease (1,2). FDG uptake in the body can be influenced by the dose of FDG injected, blood glucose level, sex, body mass index, and anesthesia, all of which can affect standard uptake value (SUV) measurements (3,4).

In veterinary medicine, anesthesia is essential during the FDG-PET scanning process because animal movement results in artifacts and spurious results (5,6). When movement occurs after injection of FDG, muscle physiological uptake of FDG increases (5,6). Therefore, FDG is usually injected under anesthesia to minimize the uptake of FDG caused by movement. However, in patients at higher risk, anesthesia can be induced immediately before the FDG-PET scan to minimize the risks associated with a longer duration of anesthesia. Although FDG-PET is followed by other procedures requiring anesthesia, such as magnetic resonance imaging, the extension of anesthesia is inevitable. In other words, the anesthetic duration for the FDG-PET scan can vary in animals according to the tolerance of the animal and the need for additional procedures.

To date, no studies have investigated changes in FDG uptake in dogs that undergo anesthesia for different durations. Therefore, this study aimed to analyze the influence of isoflurane anesthesia on FDG uptake in dogs.

Materials and methods

Animals

Six beagles (2 male and 4 female; DooYeol Biotech, Seoul, Republic of Korea) weighing 8.52 ± 1.69 kg [mean ± standard deviation (SD)] and ranging in age from 2 to 4 y (2.93 ± 0.76 y) were included in this study. All dogs were considered healthy based on physical examination, complete blood (cell) count (CBC), and serum chemistry profile. The dogs were acclimated for at least 2 wk and housed under the following conditions: relative humidity of 50% ± 10%, temperature of 20°C ± 2°C, air ventilation rate of 10 cycles/h, and 12/12 h light/dark cycle. The dogs received commercial dry food (L.I.D. Limited Ingredient Diets Potato & Duck Dry Dog Formula; Natural Balance, San Diego, California, USA) twice a day and fresh water continuously throughout the experimental period. The experimental protocol was approved by the Institutional Animal Care and Use Committee (CBNUA-1411-20) of the Laboratory Animal Research Center at Chungbuk National University.

Animal preparation and anesthesia

In a crossover trial with a 1-week washout period, each dog was anesthetized with isoflurane for 3 different durations: 60 min (Group 1), 90 min (Group 2), and 150 min (Group 3). Anesthesia was induced 30 min after, immediately before, and 60 min before FDG injection (5.18 to 6.29 MBq/kg; HDX, Seoul, Republic of Korea) in Groups 1, 2, and 3, respectively. PET images were obtained 60 min after FDG administration. All dogs fasted for 12 h before induction of anesthesia but had water ad libitum. A physical examination and a CBC were performed, and the serum chemistry profile (including serum glucose level) was assessed before induction of anesthesia and FDG administration. All dogs were injected intravenously (IV) with propofol (Provive Injection; Myungmoon, Seoul, Republic of Korea) 4 to 8 mg/kg body weight (BW) for endotracheal intubation, followed by isoflurane administration (Terrell; Piramal Critical Care, Bethlehem, Pennsylvania, USA) at 3% of the volume inspired during scanning in a circle rebreathing system. Intermittent positive pressure ventilation was applied and the tidal volume for ventilation was 10 to 20 mL/kg BW with a respiratory frequency of 10 to 15 breaths/min. A urethral catheter was inserted under sterile conditions, which included wearing sterile gloves. All dogs received 0.9% normal saline solution (sterile normal saline; Dai Han Pharm, Seoul, Republic of Korea) 5 mL/kg/h during anesthesia and were positioned in sternal recumbency within the PET/CT gantry. Vital signs, such as heart rate, oxygen saturation, end-tidal CO₂ concentration, and blood pressure, were continuously monitored. The dogs were kept warm using a warming pad (Equator Convective Warming System; SurgiVet, Saint Paul, Minnesota, USA).

¹⁸F-fluorodeoxyglucose PET/CT scanning

The PET/CT system (DiscoverySTE; GE Healthcare, Waukesha, Wisconsin, USA) has an 8-slice helical CT scanner and a cylindrical PET device with 13 440 individual crystals arranged in 24 rings. The PET scanner uses 4.7 × 6.3 × 30 mm³ crystals grouped in 8 × 6 blocks. The 24 rings of the PET system provide 47 images at 3.27-mm intervals covering an axial field of view of 157 mm. PET scans of uniform cylindrical phantoms using GE-68 were used to normalize and calibrate the scanners.

FDG (5.18 to 6.29 MBq/kg) was injected IV into a saphenous vein as a slow bolus, followed by flushing with 10 mL of 0.9% normal saline. A low-dose CT scan was performed before each PET scan. The PET images, which acquisition time required 30 min, were obtained 1 h after FDG administration. Inhalation anesthesia was maintained throughout the scanning process and FDG was injected 1 h before the initiation of the PET scan in all experimental groups.

The images were reconstructed using iterative techniques. Four iterations with 28 subsets were performed during reconstruction. The image data matrix was 256 × 256 with a pixel size of 1.95 mm and a slice thickness of 3.27 mm. Attenuation and scatter corrections were applied. A Gaussian post-reconstruction smoothing filter with a half-maximum width of 5 mm was used to achieve uniform image resolution across sites.

Image analysis

A commercial DICOM viewer (OsiriX MD v10.0; Pixmeo, Geneva, Switzerland) was used to analyze the PET images. The SUVs of the 7 gross structures (whole brain, lung, salivary gland, mediastinal blood pool, liver, spleen, and kidney cortex) and the 7 intracranial structures (frontal lobe, parietal lobe, temporal lobe, occipital lobe, cerebellum, brain stem, and caudal colliculus) were evaluated. Regions of interest (ROIs) for each organ were drawn manually in the transverse plane (lung, salivary gland, mediastinal blood pool, liver, spleen, kidney cortex, frontal lobe, parietal lobe, temporal lobe, occipital lobe, and caudal colliculus) and the midsagittal plane (whole brain, cerebellum, and brain stem) of CT images (Figures 1, 2) and then analyzed by 3 researchers (JS, YC, and BK). For semiquantitative image analysis, ROIs were drawn over 3 consecutive slices, except for the whole brain, caudal colliculus, cerebellum, and brain stem, for which ROIs were drawn over 1 slice only. The ROIs were then transferred to the corresponding region of the PET/CT fusion images and the mean and maximum SUVs [average tissue concentration of FDG (MBq/mL)/total dose injected (MBq)/BW (g)] was calculated for each ROI.

Figure 1.

Representative regions of interest (ROIs) in the 7 gross structures defined on computed tomography images. The white dotted lines indicate examples of ROIs assessed in the study. (A) whole brain, (B) lung, (C) salivary gland, (D) mediastinal blood pool, (E) liver, (F) spleen, and (G) kidney cortex.

Figure 2.

Representative regions of interest (ROIs) in the 7 intracranial structures defined on computed tomography images. White dotted lines indicate examples of ROIs assessed in the study. (A) frontal lobe, (B) parietal lobe, (C) temporal lobe, (D) occipital lobe, (E) brain stem, (F) cerebellum, and (G) caudal colliculus.

Statistical analyses

Data were analyzed using commercially available statistical software (Prism version 8; GraphPad, San Diego, California, USA). The normal distribution of the data was assessed with the Kolmogorov-Smirnov test. The test data were not normally distributed so we evaluated differences between groups using the Kruskal-Wallis test. If the differences were significant, Dunn’s multiple comparison test was used for comparisons between groups. For 150 min of isoflurane anesthesia, serum glucose level was assessed using the 1-way analysis of variance (ANOVA) test followed by Tukey’s multiple comparison test. All values in each table are expressed as the mean and the 95% confidence interval (CI). Differences were considered statistically significant at P < 0.05.

Results

All dogs were successfully examined during the FDG-PET scan and no adverse effects were observed during or after recovery from anesthesia. There were no significant differences in serum glucose concentration between the 3 groups (mean ± SD; Group 1 = 93.17 ± 7.44 mg/dL; Group 2 = 90.33 ± 9.03 mg/dL; Group 3 = 96.5 ± 9.96 mg/dL) (P = 0.77).

Gross structures

Representative ROIs in the 7 gross structures are shown in Figure 1. The mean and maximum SUVs for the 7 gross structures in the 3 groups are listed in Table I. There were no significant differences in the mean or maximum SUVs of the lung, salivary gland, mediastinal blood pool, liver, spleen, or kidney cortex between the 3 groups. On the contrary, the whole brain showed significantly lower mean and maximum SUVs in dogs with longer duration of anesthesia (P < 0.01). Post-hoc analysis of the whole brain demonstrated that Group 3 showed significantly lower mean and maximum SUVs than Group 1 (P < 0.05). There were no significant differences in the mean or maximum SUVs of the entire brain between Group 2 and Group 1 compared to Group 3.

Table I.

The mean and maximum standard uptake values (SUVs) for the 7 gross structures of 6 normal dogs evaluated with FDG positron emission tomography. Group 1 dogs were under anesthesia for 60 min, Group 2 dogs were under anesthesia for 90 min, and Group 3 dogs were under anesthesia for 150 min. The SUVs data are presented as the mean and 95% confidence interval. Superscript letters indicate groups that are significantly different based on Dunn’s multiple comparisons test.

Region	Group 1		Group 2		Group 3		P-value*
	Mean	Max	Mean	Max	Mean	Max	Mean	Max
Whole brain	3.88 (3.30 to 4.46)	5.47 (4.47 to 6.48)	2.69 (2.19 to 3.18)	4.25 (3.20 to 5.30)	2.27 (1.89 to 2.66)a	3.46 (2.77 to 4.14)b	< 0.01	< 0.01
Lung	0.38 (0.36 to 0.39)c	0.88 (0.74 to 1.02)c	0.40 (0.39 to 0.41)c	1.02 (0.70 to 1.35)c	0.39 (0.35 to 0.42)c	1.07 (0.86 to 1.27)c	0.091	0.46
Salivary gland	2.33 (1.84 to 2.81)	3.03 (2.48 to 3.58)	2.75 (2.30 to 3.21)	3.68 (2.97 to 4.39)	2.33 (1.83 to 2.83)	3.05 (2.41 to 3.68)	0.22	0.21
Mediastinal blood pool	1.30 (1.83 to 2.83)d	1.62 (1.42 to 1.82)d	1.43 (1.13 to 1.74)	1.80 (1.50 to 2.09)	1.30 (1.04 to 1.56)	1.65 (1.37 to 1.94)d	0.61	0.53
Liver	1.38 (1.06 to 1.70)d	1.59 (1.26 to 1.93)c	1.58 (1.24 to 1.92)	1.84 (1.45 to 2.23)	1.43 (1.18 to 1.67)	1.69 (1.41 to 1.97)d	0.55	0.51
Spleen	0.91 (0.76 to 1.05)c	1.40 (1.14 to 1.66)c	1.07 (0.85 to 1.28)c	1.51 (0.93 to 2.10)c	0.98 (0.83 to 1.13)c	1.52 (1.11 to 1.92)d	0.31	0.98
Kidney cortex	1.73 (1.11 to 2.35)	3.71 (2.46 to 4.96)	1.82 (1.46 to 2.17)	3.35 (2.63 to 4.08)	1.69 (1.36 to 2.01)	3.32 (2.51 to 4.13)	0.51	0.88
P-value†	< 0.001	< 0.001	< 0.001	< 0.001	< 0.001	< 0.001

^*Comparison among the 3 groups.

^†Comparison among the 7 gross structures.

^aDifference between Group 1 and Group 3 (P < 0.01).

^bDifference between Group 1 and Group 3 (P < 0.05).

^cDifference compared to the whole brain (P < 0.01).

^dDifference compared to the whole brain (P < 0.05).

In each group, the mean and maximum SUVs were significantly different between the ROIs (P < 0.01). In particular, the highest uptake was observed in the whole brain, followed by the salivary gland and the kidney cortex. The lowest uptake was observed in the lung, followed by the spleen and the mediastinal blood pool.

Intracranial structures

Representative ROIs of the different brain regions are shown in Figure 2. Representative PET/CT fusion images of the different regions of the brain are shown in Figure 3. Table II shows the mean and maximum SUVs of the intracranial structures in the 3 groups. In the brain, the mean and maximum SUVs of the frontal lobe, parietal lobe, temporal lobe, occipital lobe, cerebellum, brain stem, and caudal colliculus were significantly different between the 3 groups (P < 0.05). Post-hoc analysis revealed that Group 3 had significantly lower mean and maximum SUVs than Group 1 in all regions of the brain (P < 0.05). There were no significant differences in the mean SUVs of the intracranial structures between Group 2 and Group 1 compared to Group 3. In each group, the highest uptake was observed in the caudal colliculus, followed by the cerebellum. In contrast, the lowest uptake was reported in the occipital lobe.

Figure 3.

Representative positron emission tomography (PET)/computed tomography (CT) fusion images at the level of the frontal lobe (A), parietal lobe, temporal lobe (B), caudal colliculus (C), occipital lobe (D), and brainstem and cerebellum (E). PET/CT images are scaled identically between 0 and 10, with respect to standardized uptake value. Group 1 — dogs under anesthesia for 60 min. Group 2 — dogs under anesthesia for 90 min. Group 3 — dogs under anesthesia for 150 min.

Table II.

The mean and maximum standard uptake values (SUVs) for the 7 intracranial structures of 6 normal dogs evaluated with FDG positron emission tomography. Group 1 dogs were under anesthesia for 60 min, Group 2 dogs were under anesthesia for 90 min, and Group 3 dogs were under anesthesia for 150 min. The SUVs data are presented as the mean and 95% confidence interval (CI).

Region	Group 1		Group 2		Group 3		P-value*
	Mean	Max	Mean	Max	Mean	Max	Mean	Max
Frontal lobe	3.77 (3.03 to 4.51)	4.65 (3.79 to 5.52)	2.62 (2.26 to 2.98)	3.34 (2.74 to 3.93)	2.22 (1.79 to 2.65)a	2.80 (2.25 to 3.35)a	< 0.01	< 0.01
Parietal lobe	3.87 (3.21 to 4.53)	5.25 (4.39 to 6.10)	2.45 (2.02 to 2.88)	3.05 (2.54 to 3.57)	2.09 (1.71 to 2.48)a	2.70 (2.26 to 3.14)a	< 0.01	< 0.01
Temporal lobe	3.30 (2.80 to 3.81)	4.48 (3.65 to 5.31)	2.48 (2.04 to 2.92)	3.02 (2.46 to 3.59)	2.13 (1.68 to 2.58)a	2.57 (1.97 to 3.16)a	< 0.01	< 0.01
Occipital lobe	2.72 (2.21 to 3.22)	3.83 (3.38 to 4.28)	1.91 (1.50 to 2.31)	2.59 (2.23 to 2.95)	1.53 (1.13 to 1.94)a	2.16 (1.74 to 2.58)a	< 0.01	< 0.001
Cerebellum	4.38 (3.39 to 5.36)	5.37 (4.38 to 6.36)	3.55 (2.78 to 4.32)	4.25 (3.20 to 5.30)	2.86 (2.40 to 3.33)b	3.45 (2.77 to 4.14)a	0.016	< 0.01
Brain stem	3.08 (2.52 to 3.65)	4.41 (3.62 to 5.19)	2.62 (2.25 to 3.00)	3.52 (2.84 to 4.20)	2.15 (1.72 to 2.58)b	2.98 (2.47 to 3.49)b	0.014	< 0.01
Caudal colliculus	4.99 (4.09 to 5.90)	5.48 (4.41 to 6.56)	4.03 (3.15 to 4.91)	4.37 (3.46 to 5.29)	3.20 (2.43 to 3.96)b	3.52 (2.70 to 4.33)b	< 0.01	< 0.01

^*Comparison among the 3 groups.

^aDifference between Group 1 and Group 3 (P < 0.01).

^bDifference between Group 1 and Group 3 (P < 0.05).

Discussion

The purpose of this study was to evaluate the influence of isoflurane anesthesia on FDG uptake in dogs. We determined that as the length of time under isoflurane anesthesia increased, FDG uptake decreased in the brain. However, in other organs, there were no significant associations between anesthesia duration and FDG uptake.

Several studies have shown that anesthesia impedes FDG uptake in the brain (3,4,7,8). The effect of isoflurane anesthesia on cerebral glucose metabolism has been demonstrated in humans (4). Specifically, cerebral glucose metabolism decreases uniformly by 46% with isoflurane anesthesia in humans (4). In the present study, the whole brain and the 7 intracranial structures showed significantly lower mean and maximum SUVs in dogs with prolonged anesthesia, which corresponds to the results of previous studies. Specifically, there were significant differences between Groups 1 and 3 in all areas of the brain and brain FDG uptake in Group 3 was approximately 38.1% ± 5.56% less than in Group 1.

The brain increases glucose metabolism with activation of neural tissues (9) and cerebral glucose metabolism measured by FDG is an indirect measure of neuronal function. The mechanism of FDG uptake is based on transport through the glucose transporter (GLUT) family and glucose phosphorylation by hexokinase (9,10). Volatile anesthetics affect neuronal ion channels, particularly fast-acting synaptic neurotransmitter receptors such as gamma-aminobutyric acid (GABA), nicotinic acetylcholine, and glutamate receptors (11–13). Isoflurane anesthesia increases the sensitivity of GABA receptors and prolongs the duration of GABA-mediated synaptic inhibition (14). An increase in GABA concentration under isoflurane was observed using magnetic resonance spectroscopy in mice (15). Thus, a decrease in FDG uptake by the brain, unlike other organs, could be explained by the depleted neuronal activity caused by isoflurane.

Anesthetic agents, including inhalation anesthetics, solubilize active mitochondrial hexokinase in a less active soluble form in the brain, inhibiting glucose phosphorylation (16,17). The ratio of the affinities of brain hexokinase for FDG and glucose is decreased under isoflurane anesthesia (18). Therefore, inhibited hexokinase activity could be one reason for the decreased uptake of FDG identified in Group 3.

Glucose and FDG enter the cell by the same GLUT that can be saturated. Therefore, increased glycemia leads to competitive saturation of GLUT and may decrease FDG uptake. Within the normoglycemic range, elevated glucose levels interfere with FDG uptake by the brain (19). In humans, an increase in glycemia from 60 to 120 mg/dL leads to an almost 50% reduction in FDG uptake by the brain due to the limited transport mechanism of GLUT (20). In this study, Group 3, which had the lowest brain SUVs, was injected with FDG after 1 h of isoflurane anesthesia and was the most affected by isoflurane. Previous studies have reported that serum glucose level increases with anesthetic agents, including isoflurane (21–23). Increased glycemia during isoflurane anesthesia has been reported in dogs (21). Volatile anesthetics induce an increase in blood glucose levels due to impaired glucose-induced insulin release. Isoflurane inhibits insulin release by activation of the sarcolemmal K_ATP channel (24) and elevated plasma glucose levels interfere with cerebral FDG uptake (25–27). Furthermore, insulin-independent GLUTs (i.e., GLUT1 and GLUT3) are prevalent.

In the brain, reduced insulin release does not have a direct effect on GLUT activity. Serum glucose could not be monitored during anesthesia in this study to minimize radiation exposure. Instead, the same 6 dogs were anesthetized with isoflurane without FDG injection. Anesthesia was performed for 150 min in the same way and the serum glucose level of the 6 dogs was measured before induction of anesthesia and every 30 min thereafter (mean ± SD; 0 min = 102.50 ± 15.57 mg/dL; 30 min = 101.83 ± 28.29 mg/dL; 60 min = 138.17 ± 18.17 mg/dL; 90 min = 132.83 ± 19.14 mg/dL; 120 min = 144.00 ± 28.95 mg/dL; 150 min = 137.00 ± 18.02 mg/dL). Similar to the previous study (21), an elevation of serum glucose was observed from induction to 150 min of isoflurane anesthesia.

One concern is whether an increased plasma glucose level affects the uptake of FDGs from other organs. The serum glucose level has a significant influence on the uptake of FDG. However, many reports have suggested that the brain is the only organ in which hyperglycemia significantly affects SUVs (28–30). The effect of glycemia on FDG uptake is pronounced in the brain, whereas organs such as the liver and spleen that consume less glucose show an insignificant effect (28). This pattern supports the results that there were no significant differences in the mean SUVs of the organs other than the brain region. FDG-PET is a semi-quantitative method, providing SUVs rather than kinetic modeling results (31–33). To explain the uptake of FDG in organs other that the brain, it is necessary to study how isoflurane anesthesia alters the affinities of FDG versus glucose with GLUT and hexokinases as well as the volume distribution.

Anesthesia increases the ratio of brain to serum glucose whether or not serum glucose is increased (34,35). Increases in extracellular glucose in the brain have previously been reported with various anesthetics, including isoflurane (36,37). The decreased glucose consumption by brain cells and the vasodilation of cerebral vessels could contribute to the development of brain hyperglycemia. However, the underlying mechanisms of brain hyperglycemia caused by anesthesia remain a matter of speculation (38). Increased extracellular glucose in the brain may decrease cerebral FDG uptake based on competitive inhibition of glucose in FDG.

In a previous study, dogs were kept in a room after injection of FDG for 1 h and anesthetized with isoflurane during a 20-minute PET scan period (39). The duration of anesthesia used in the previous study was shorter than in this study. In our study, the mean SUVs of the brain region in Groups 2 and 3 were less than those of the previous study. Specifically, we determined that the mean SUVs of the brain regions decreased with prolonged anesthesia.

Depressed cardiac output can decrease the metabolic rate, which ultimately reduces FDG uptake. Cardiovascular function can also be reduced under anesthesia (40). However, the dogs maintained their vital signs (e.g., heart rate, blood pressure, body temperature, etc.) within normal ranges during the entire anesthesia period. In addition, we monitored electrocardiography during the anesthetic period but there was no problematic change. Thus, we could not confirm the decreased uptake of FDG due to decreased cardiovascular function during anesthesia.

Our study has several limitations. First, the data were obtained from clinically healthy beagle dogs. The results of oncologic patients can be different from these results because many tumors have a much higher metabolism than normal cells. Second, the effect of glucose and insulin could not be verified as serum glucose and insulin concentrations were not monitored during the FDG-PET/CT acquisition. FDG-PET is an indirect method for measuring glucose metabolism. It is a semiquantitative method that only provides SUVs rather than the kinetics of glucose metabolism. Isoflurane can alter the affinities of GLUT, hexokinase, and volume distribution. More studies are needed to examine alterations in GLUT, hexokinase, and volume distribution due to isoflurane anesthesia.

In conclusion, this study identified changes in FDG uptake depending on different anesthesia durations. Significant decreases in SUVs in brain regions were observed with increased durations of anesthesia. Therefore, veterinarians performing PET/CT should consider that the duration of anesthesia could alter the uptake of FDG in brain regions.