Quantification of metabolic rate of glucose uptake with [18F]FDG
[18F]2-fluoro-2-deoxy-D-glucose ([18F]FDG) is a glucose analogue, where fluorine-18 (halflife 109.8 min) substitutes the hydroxyl group at the second position in the glucose molecule. [18F]FDG is commonly used to measure tissue glucose consumption in vivo.
[18F]FDG enters the tissue via glucose transporters, and can then be either phosphorylated by hexokinases to [18F]FDG-6-phosphate ([18F]FDG-6-P), or transported from tissue back to blood. [18F]FDG-6-P cannot be transported out of the tissue, and is not metabolized further through glycolysis like glucose-6-phosphate. In most tissues dephosphorylation of [18F]FDG-6-phosphate is slow because of very low activity of glucose-6P-phosphatase. Depending on the tissue and animal species, further metabolism of [18F]FDG-6-P, especially though pentose phosphate pathway (PPP), may be significant (Kanazawa et al., 1996 and 1997; Southworth et al., 2003; Herrero et al., 2004; Rokka et al., 2017; Klebermass et al., 2021). These metabolites too are mainly trapped inside cells. Within endoplasmic reticulum (ER), hexose-6-phosphate dehydrogenase (H6PD) can dehydrogenate [18F]FDG-6-phosphate, enhancing its trapping, and H6PD inhibitor metformin can reduce the retention of [18F]FDG (Marini et al., 2016; Cossu et al., 2019).
[18F]FDG does not distribute into the intracellular lipid droplets, which will decrease the apparent tracer uptake measured per tissue volume; this should be considered when studying organs with high and variable fat content.
[18F]FDG study may be combined with hyperinsulinemic euglycemic clamp (DeFronzo et al., 1979) to assess the insulin sensitivity (or insulin resistance) of specific organs (Nuutila et al., 1993; Johansson et al., 2017). The whole body insulin sensitivity can be measured simultaneously as the M value by dividing the mean glucose infusion rate by the lean body mass.
[18F]2-fluoro-2-deoxy-D-glucose has high affinity for GLUTs, and low affinity for ATP-dependent sodium-glucose transporters SGLTs. The relative role of GLUTs and SGLTs can be studied with related glucose analogue tracers (Sala-Rabanal et al., 2016; Barrio et al., 2020).
[18F]FDG model
The autoradiographic method for measuring regional metabolic rate of glucose in the brain of rat using [14C]deoxyglucose (Sokoloff et al. 1977) has been modified for human studies using positron emission tomography (PET) and [18F]2-fluoro-2-deoxy-D-glucose (Phelps et al., 1979; Reivich et al., 1979; Huang et al, 1980 and 1981).
The three-compartment model with four rate constants K1*, k2*, k3* and k4* is often simplified by assuming that the dephosphorylation rate of FDG-6-phosphate in brain tissue is small enough that it can be ignored (k4*=0). At least a part of the observed k4* may be explained by tissue heterogeneity (Schmidt et al., 1992). In the brain, BBB is the first rate-limiting step of [18F]FDG, as implicated by even distribution of glucose in the brain intra- and extracellular spaces (Pfeuffer et al., 2000), and therefore rate constants K1* and k2* in the brain represent the transport across BBB, and k3* represents the rate of intracellular phosphorylation. This is the case in the brain, not in other tissues, where the endothelial wall is highly permeable to glucose: K1* may then represent either (plasma) perfusion, transport from extracellular into intracellular space, or combination of those. The initial uptake phase of [18F]FDG may then be used as an estimate of perfusion (Mullani et al., 2008; Zhang et al., 2023). With the increased sensitivity of PET scanners with long axis field of view, high temporal resolution can be achieved, and a 90 s PET scan can provide whole body maps of K1 for [18F]FDG and time delay (Feng et al., 2021).
Metabolic rate of glucose (MRglu) can be calculated from equation
, where Cglu is the concentration of glucose in plasma, and LC is the lumped constant.
Originally FDG studies were always analysed using compartment model with the three or four rate constants. Only later it was found out that a graphical method, Patlak plot can be used to directly estimate the combination of the model rate constants. This combined term, net uptake rate for FDG (Ki*) is robust and very fast to calculate, and is therefore also suitable for computation of parametric images.
Patlak plot provides not only Ki*, but also an index of FDG distribution volume in the tissue as the plots intercept with y axis. In tissues with high air or fat content both the distribution volume and Ki* will be reduced. Metabolic rate of glucose can be corrected for this effect by dividing Ki* with the Patlak plot intercept value (Jones et al., 1997; Keramida et al., 2016).
Semiquantitative methods (for example SUV), which do not require blood sampling, are be affected by differences in plasma clearance, that is, uptake to other organs, or excretion of FDG: for example, medication for hypertension may slow down the renal excretion of FDG (Zhao et al., 2013). Akers et al (2016) did not find any correlation between FDG SUV and renal function (eGFR). Correction for plasma glucose concentration, in calculation of glucose consumption is straightforward for Ki* but not for SUV.
For estimation of metabolic rate of glucose, if dynamic PET data is collected, then the Patlak graphical analysis is recommended for calculation of Ki , and from that, the metabolic rate. The Ki can be calculated from regional TACs or as parametric Ki image. If only static PET scan is available, then Patlak plot can not be used; calculate FUR as a substitute for Ki , either from regional TACs or as parametric FUR image.
Input function for FDG
Since all living tissues utilize glucose, there is no reference tissue that could be used for quantitative analysis of FDG PET data instead of plasma input function.
Manual blood samples are collected for the measurement of concentration of FDG in plasma, to enable absolute quantification of glucose uptake. Arterialized venous blood sampling is often used instead of arterial sampling in FDG studies, although it increases the variability and larger sample size is needed. Blood samples are processed in the PET blood laboratory to plasma time-activity curves, which can be used as such. Image-derived input function (van der Weerdt et al., 2001; Christensen et al., 2014), Model-based input function, or population-based input function may be used as an alternative to blood sampling in some cases. When necessary, blood TAC can be converted to plasma TAC, or vice versa, using conversion functions (Figure 2) that are available in programs b2plasma and p2blood. In humans, FDG is rapidly equilibrated between plasma and RBC (Gambhir et al., 1989; Nahmias et al., 2000), although not to exactly same level (Ohtake et al., 1991). The blood-to-plasma ratio is ∼0.9, except in the very beginning, and using arterial blood instead of plasma overestimates Patlak Ki only ∼5% (Larsen et al., 2001). On the other hand, RBCs can transfer glucose into human brain in such quantities that it affects brain function (Wang et al., 2023), and therefore plasma curve may not actually be better input function than blood curve in human FDG studies. Note that FDG transport across erythrocyte membranes is usually slow in nonprimates, and therefore plasma TAC should be preferred over blood TAC as input function (Alf et al., 2013; Buxton, 2014). Fetal and neonatal erythrocytes of all mammalian species are highly permeable to glucose and FDG. Metabolites of FDG (Herrero et al., 2004) may have some effect on the plasma-to-blood ratio.
Figure 2. Plasma-to-blood ratio functions for FDG in human subjects, rats, and mice. The different conversion functions for mice, based on separate publications, may indicate differences between mice populations.
For Patlak analysis, correction for time delay is not required, but the estimates of compartmental model parameters may be very sensitive to the effect of time delay.
Brain
With high-resolution PET scanner (HRRT) and careful optimization of image reconstruction it is possible to avoid arterial blood sampling by deriving blood TAC from carotid artery in the dynamic image (Huisman et al., 2012). Three manual blood samples are needed to calibrate the image-derived input curve.
Alternatively, in diagnostic brain studies the other hemisphere may be used to calculate asymmetry index. Disease models in animals are often designed to, for instance, use the other brain hemisphere or cerebellum as healthy reference. In clinical FDG PET studies of Alzheimer's disease, assessment is usually based on "metabolic ratio", using a set of well-preserved brain regions such as pons or cerebellar vermis as reference region.
In brain activation studies, PET image is often normalized using whole brain or cerebellum as reference (Clark et al., 1985; Dukart et al., 2010; Hua et al., 2015), or reference is extracted using clustering (Yakushev et al., 2009; Borghammer et al., 2009). In bolus FDG method the functional metabolic responses to stimuli are presumed to sustain a constant state. Instead of bolus administration, constant FDG infusion protocols can be used to measure task-specific glucose metabolism changes (fPET) in a single scan. With FDG bolus administration protocol separate baseline and task-specific scans are required, which is hampered by high inter-individual variance (Hahn et al., 2016).
Compartment model fit for the brain can be calculated using fitk3 or fitk4. Parametric images of individual rate constants can be computed using linearized model or basis function method (Hong & Fryer, 2010), and used in SPM and connectivity analyses (Bahri et al., 2018).
LC is different in irreversible and reversible model. If LC at different study conditions is to be estimated, too, then use fcmrglu, for example:
fcmrglu ua2826ap.delay.kbq ua2826dy1.dft 5.2 ua2826fcmrglu.res
Make sure that you are using appropriate values for τ and φ. Note that the brain FDG model must not be applied to skeletal muscle!
Myocardium
Please read MET5731. Recommended LC for Patlak analysis of heart FDG studies is 1. If myocardial FDG data is analyzed using irreversible compartmental model, individual LC can be calculated based on the model rate constants (Bøtker et al., 1997 and 2000).
For precise quantification the analysis methods should account for the spillover and partial volume effects caused by respiratory motion and beating of the heart, especially in small animal studies. Gated PET images help to reduce these artifacts, and may additionally enable analysis of cardiac motion in CAD (Sims et al., 2018).
In heart studies the input function is usually derived from a ROI placed on the LV cavity or aorta. Arterialized venous blood sampling can be used instead of arterial sampling, although it contributes to variability in results (van der Weerdt et al., 2002).
Compartmental model for FDG provides K1, which, after normalized for plasma glucose level and extraction fraction correction, could provide estimates of myocardial perfusion. (Zuo et al., 2021).
Skeletal muscle
For analysis instructions in TPC, please read MET5736, and the publication by Bertoldo et al. (2001). Recommended LC for Patlak analysis of skeletal muscle FDG studies is 1.2 (Peltoniemi et al., 2000). If compartmental analysis is used, it should include three tissue compartments (Bertoldo et al., 2001; Huang et al., 2011; Bahri et al., 2018).
From the femoral region, image-derived input function can be obtained, as validated by Christensen et al (2014).
In small animal studies the data quality does not usually allow fitting complex compartmental models. 2-tissue compartmental model for irreversible uptake has been applied to gastrocnemius muscles of mice, using image-derived input function from inferior vena cava (Cochran et al, 2016).
Liver
Hepatic glucose uptake can be estimated using FDG, if the dual input (arterial and portal vein input) is taken into consideration in the model, as validated in pig studies (Brix et al., 2001; Munk et al., 2001; Iozzo et al., 2007; Kudomi et al., 2009; Winterdahl et al., 2011; Rani et al., 2013). Alternatively the gut could be included as a compartment in the model (Vivaldi et al., 2013; Garbarino et al., 2015). Constant-infusion administration of FDG may yield more robust estimates of hepatic glucose metabolism than bolus injection (Trägårdh et al., 2015).
Notice that liver and intestine are gluconeogenetic organs, expressing glucose-6-phosphatase, and therefore we cannot assume that FDG uptake is irreversible. Yet, desphophorylation in the liver is relatively slow compared with phosphorylation, even during fasting, but especially during euglycemic hyperinsulinemic clamp (Iozzo et al., 2004; Rijzewijk et al., 2010). Therefore, Patlak plot (assuming irreversible uptake) is often used to analyse FDG liver data. Also Logan plot, assuming reversible uptake and providing distribution volume of the tracer, has been applied (Lauritsen et al., 2020).
Arterial plasma curve alone can be used as model input especially for liver tumours which are mainly fed by the hepatic artery (Choi et al., 1994; Keiding et al., 2000; Brix et al., 2001; Fukuda et al., 2004). Image-derived input function from abdominal aorta can be used in the analysis. Calculation of liver-to-blood ratio is suitable method for clinical studies, and provides better surrogate for Ki than SUV.
Hepatic glucose uptake is inversely associated with liver fat content (Borra et al., 2008; Rijzewijk et al., 2010). Glucose uptake in liver may be higher in men than in women, even when corrected for the different fat contents (Keranida and Peters, 2017).
Liver as reference region
Perfusion in liver is high, and blood-to-hepatocyte transfer is practically unlimited for [18F]FDG; therefore radioactivity concentration in the liver follows relatively well the [18F]FDG concentration in the blood. Liver is often used as a reference tissue in diagnostic imaging when calculating SUV ratio (SUR, or tumour-liver ratio, TLR) (Laffon et al., 2011; Boktor et al., 2013; Watanabe et al., 2013; Keramida et al., 2015; Hofheinz et al., 2016) or metabolic volume. TLR (or Deauville score) has high prognostic value for instance in imaging of lymphoma (Barrington et al., 2017; Nanni et al., 2017). Liver has also been used as reference tissue in mice BAT studies after intraperitoneal FDG injection (Wu et al., 2014).
SUV in liver correlates positively with blood glucose levels, even when patients have been fasting and serum glucose concentrations are considered to be at acceptable levels (Webb et al., 2015), has high inter-individual variance and is dependent on BMI and sex (Rubello et al., 2015). In contrast to liver, tumour SUV correlates negatively with blood glucose, making TLR very sensitive to blood glucose levels (Peters et al., 2018). Although tumour-to-blood ratio is preferable to tumour-to-liver ratio, both are clearly better than SUV in oncological FDG PET (van den Hoff et al., 2013; Hofheinz et al., 2016).
Liver SUV is decreased in steatotic liver, which causes overestimation of tumour-to-liver ratio or Deauville score, unless SUV is corrected using liver CT-derived HU-values (Salomon, 2018). HU-corrected liver SUV is increased in hepatic steatosis (Keramida et al., 2014). Fat distribution in the liver can be very heterogeneous (Décarier et al., 2011; Keramida et al., 2016).
Intestine
Metabolic rate of glucose in intestine (duodenum and jejunum) and colon in humans has been measured using [18F]FDG PET (Honka et al., 2013; Mäkinen et al., 2015; Bahler et al., 2017; Kang et al., 2017; Koffert et al., 2017; Motiani et al., 2017). Honka et al (2013) and Motiani et al (2017) applied LC=1.15, and a recovery coefficient of 2.5 for the colon. Intravenously administered [18F]FDG reflects intestinal tissue accumulation (mainly in mucosal layer), with no uptake in intestinal lumen (Honka et al., 2013).
[18F]FDG PET has also been used to assess glucose metabolism in the gut of mice (Massollo et al., 2013). Sala-Rabanal et al (2018) have studied intestinal absorption of glucose in mice using FDG and two other tracers, [18F]Me-4FDG and [18F]4-FDG, with different specificities for SGLTs and GLUTs.
Lungs
Lung tumours and infections and inflammatory diseases are commonly studied using [18F]FDG PET. Quantitation of glucose metabolism in normal lung requires input function from pulmonary aorta or RV cavity, and a lung-specific model (Schroeder et al., 2008; Chen et al., 2017). Lung SUV and Ki must be corrected for the tissue fraction, which can be calculated from transmission scan or CT derived HU values (Lambrou et al., 2011; Holman et al., 2015). Patlak plot Ki can be normalized for differences in lung tissue density (air content) by dividing it by the y axis intercept (Jones et al., 1997).
Bone
[18F]FDG can be used to study bone infections and to separate normal bone healing from bone infections (Koort et al. 2004 and 2005; Lankinen et al., 2012; Odekerken et al., 2014).
Bone marrow
Metabolism of glucose in bone marrow can be measured using [18F]FDG PET (Huovinen et al., 2014; Latva-Rasku et al., 2018). [18F]FDG PET can be used to detect bone marrow metastases in an early phase (Evangelista et al., 2012), and for assessment of multiple myeloma (van Lammeren-Venema et al., 2012; Paschali et al., 2021). FDG uptake has been used as an index of functional bone marrow during radio- and chemotherapy of cancer (Elicin et al., 2014; Yagi et al., 2015; Abravan et al., 2019).
Cartilaginous tumours can be graded using [18F]FDG PET-CT (Annovazzi et al., 2019).
Tumours
Glucose consumption in tumours is increased because of increased need for building blocks and energy (Warburg effect). GLUT1 and GLUT3 are commonly overexpressed by tumour cells, and the level of hexokinase 2 is usually high. The enhanced glucose uptake can be quantified or semi-quantified using [18F]FDG PET. Part of the increase may be due to inflammation. [18F]FDG is not well suitable for imaging all tumours: neuroendocrine and prostate tumours often have low [18F]FDG uptake. Increased glucose transport may be due to SGLTs which do not transport [18F]FDG. Certain tumour cells may have low hexokinase 2 expression or increased glucose-6-phosphatase expression.
Clinical oncology studies should follow the published recommendations and guidelines (Shankar et al., 2006; Boellaard et al., 2015), even though it may be difficult to do so (de Jong et al., 2017).
Adipose tissue
[18F]FDG can be used to assess the glucose uptake in white adipose tissue (WAT). FDG PET studies have shown that adipose tissue becomes insulin resistant in obesity and type 2 diabetes (Virtanen et al., 2001; Virtanen et al., 2005). Lumped constant for white adipose tissue is 1.14 (Virtanen et al., 2001). In type 2 diabetics, FDG SUV correlates negatively with insulin resistance and positively with adiponectin plasma concentration (Reijrink et al., 2021).
Activity of brown adipose tissue (BAT) can be measured using FDG PET (Virtanen et al., 2009). FDG is also useful in identifying and locating the BAT for defining regions-of-interest in PET image analysis when using other tracers with lesser target-to-background contrast. FDG uptake in BAT is higher when fat fraction in BAT is smaller (Koskensalo et al., 2017; Lundström et al., 2021), which may be caused by the subsequently higher water fraction and distribution volume.
Kidneys
Glucose is completely reabsorbed in the proximal tubules of the kidneys, but FDG is only partially reabsorbed (Kosuda et al., 1997; Moran et al., 1999). Reabsorption of 2-deoxy-D-glucose and FDG happens in distal tubules, possibly via GLUT1 (Miller et al., 1992; Kosuda et al., 1997). Partial excretion of FDG leaves high radioactivity concentration in the urine, hampering PET imaging of the pelvic regions, especially the kidneys and urinary bladder, but also prostate and uterine. Urinary bladder receives the highest radiation dose from FDG scan. Renal failure and medications can alter FDG excretion, which affects the total clearance of FDG and may therefore cause variability in SUV measures of other organs (Minamimoto et al., 2007; Bach-Gansmo et al., 2012; Garbarino et al., 2014). Yet, Akers et al (2016) and Kode et al (2017) did not find any correlation between FDG SUV and renal function (eGFR). In hemodialysis patients the SUV values are markedly increased (Toriihara et al., 2015).
Normal kidneys do not store appreciable amounts of glycogen (Biava et al., 1966), but can produce and release significant amounts of glucose via gluconeogenesis from circulating lactate, glycerol, and glutamine (Stumvoll et al., 1997). Gluconeogenesis is restricted to proximal tubules in the renal cortex, where the activities of related enzymes, including glucose-6-phosphatase, are high, while the activities of glycolytic enzymes are low in the cortex and high in the renal medulla (Burch et al., 1978; Guder & Ross, 1984). Therefore the uptake and phosphorylation of [18F]FDG in the renal cortex is reversible, like in the liver. FDG uptake in medulla is higher than in cortex (Reuter et al., 2009; Rebelos et al., 2023). FDG uptake in cortex, but not in medulla, is higher during insulin clamp than in fasting state (Rebelos et al., 2023).
FDG may have a role in diagnosing and staging primary and metastatic renal cell carcinoma (Krishnan et al., 2017), and renal inflammation (Wan et al., 2018). In clinical FDG whole-body PET, renal uptake is usually comparable to that in the liver. High renal FDG uptake has been seen in interstitial nephritis (Katagiri et al., 2010), glomerulonephritis (Holt et al., 2018), sarcoidosis (Toyonaga et al., 2014), IgG4-related disease (Bélissant et al., 2015), and AKI (Kidera et al., 2022).
Dynamic FDG PET may provide similar measures of kidney function as MAG3 scintigraphy (Geist et al., 2018 and 2019).
Endogenous glucose production
[18F]FDG study during hyperinsulinemic euglycemic clamp can provide information necessary to measure endogenous glucose production (EGP) (Iozzo et al., 2006); FDG plasma clearance is adjusted with the FDG lost in urine to determine the rate of glucose disappearance, and EGP is calculated by subtracting glucose infusion rate from the rate of disappearance.
See also:
- Net influx rate
- Multiple-time graphical analysis (MTGA)
- Fitting compartmental models
- SUV
- Tissue-to-plasma ratio
- Fractional uptake rate (FUR)
- Metabolic tumour volume (MTV)
- fPET FDG
Literature
Barrio JR, Huang S-C, Satyamurthy N, Scafoglio CS, Yu AS, Alavi A, Krohn KA. Does 2-FDG PET accurately reflect quantitative in vivo glucose utilization? J Nucl Med. 2020; 61(6): 931-937. doi: 10.2967/jnumed.119.237446.
Basu S, Zaidi H, Holm S, Alavi A. Quantitative techniques in PET-CT imaging. Curr Med Imaging Rev. 2011; 7: 216-233. doi: 10.2174/157340511796411186.
Bertoldo A, Peltoniemi P, Oikonen V, Knuuti J, Nuutila P, Cobelli C. Kinetic modeling of [18F]FDG in skeletal muscle by PET: a four-compartment five-rate-constant model. Am J Physiol Endocrinol Metab. 2001; 281:E524-E536. doi: 10.1152/ajpendo.2001.281.3.E524.
Boellaard R, Delgado-Bolton R, Oyen WJ, Giammarile F, Tatsch K, Eschner W, Verzijlbergen FJ, Barrington SF, Pike LC, Weber WA, Stroobants S, Delbeke D, Donohoe KJ, Holbrook S, Graham MM, Testanera G, Hoekstra OS, Zijlstra J, Visser E, Hoekstra CJ, Pruim J, Willemsen A, Arends B, Kotzerke J, Bockisch A, Beyer T, Chiti A, Krause BJ. FDG PET/CT: EANM procedure guidelines for tumour imaging: version 2.0. Eur J Nucl Med Mol Imaging 2015; 42(2): 328-354. doi: 10.1007/s00259-014-2961-x.
Dai X, Chen Z, Tian J. Performance evaluation of kinetic parameter estimation methods in dynamic FDG-PET studies. Nucl Med Commun. 2011; 32(1): 4-16. doi: 10.1097/MNM.0b013e32833f6c05.
DeFronzo RA, Tobin JD, Andres R. Glucose clamp technique: a method for quantifying insulin secretion and resistance. Am J Physiol. 1979; 237(3): E214-E223. doi: 10.1152/ajpendo.1979.237.3.E214.
Gjedde A. Positron emission tomography of brain glucose metabolism with [18F]fluorodeoxyglucose in humans. In: Hirrlinger J, Waagepetersen HS (eds.), Brain Energy Metabolism, Neuromethods, vol 90, Springer, 2014, p 341-364. doi: 10.1007/978-1-4939-1059-5_14.
Guedj E, et al. EANM procedure guidelines for PET brain imaging using [18F]FDG, version 3. Eur J Nucl Med Mol Imaging 2022; 49: 632-651. doi: 10.1007/s00259-021-05603-w.
Hawkins RA, Phelps ME, Huang S-C, Kuhl DE. Effect of ischemia on quantification of local cerebral glucose metabolic rate in man. J Cereb Blood Flow Metab 1981; 1: 37-51. doi: 10.1038/jcbfm.1981.5.
Honka H, Hannukainen JC, Tarkia M, Karlsson H, Saunavaara V, Salminen P, Soinio M, Mikkola K, Kudomi N, Oikonen V, Haaparanta-Solin M, Roivainen A, Parkkola R, Iozzo P, Nuutila P. Pancreatic metabolism, blood flow, and β-cell function in obese humans. J Clin Endocrinol Metab. 2014; 99(6): E981-E990. doi: 10.1210/jc.2013-4369.
Huang S-C, Phelps ME, Hoffman EJ, Sideris K, Selin CJ, Kuhl DE. Noninvasive determination of local cerebral metabolic rate of glucose in man. Am J Physiol. 1980; 238: E69-E82. doi: 10.1152/ajpendo.1980.238.1.E69.
Huang X, Bao S, Huang S-C. Clustering-based linear least square fitting method for generation of parametric images in dynamic FDG PET studies. Int J Biomed Imaging. 2007: 65641. doi: 10.1155/2007/65641.
Kalliokoski KK, Bojsen-Møller J, Seppänen M, Johansson J, Kjaer M, Teräs M, Magnusson SP. Contraction-induced [18F]-fluoro-deoxy-glucose uptake can be measured in human calf muscle using high-resolution PET. Clin Physiol Functional Imag. 2007; 27(4): 239-241. doi: 10.1111/j.1475-097X.2007.00744.x.
Larsen H, Munk OL, Keiding S. Blood FDG measurements can replace plasma FDG measurements in dynamic positron emission tomography (PET) studies. Clin Physiol Funct Imaging 2001; 21(4): 402.
Nuutila P, Knuuti J, Ruotsalainen U, Koivisto VA, Eronen E, Teräs M, Bergman J, Haaparanta M, Voipio-Pulkki L-M, Viikari J, Rönnemaa T, Wegelius U, Yki-Järvinen H. Insulin resistance is localized to skeletal but not heart muscle in type 1 diabetes. Am J Physiol. 1993; 264: E756-E762. doi: 10.1152/ajpendo.1993.264.5.E756.
Nuutila P, Peltoniemi P, Oikonen V, Larmola K, Kemppainen J, Takala T, Sipilä H, Oksanen A, Ruotsalainen U, Bolli GB, Yki-Järvinen H. Enhanced stimulation of glucose uptake by insulin increases exercise-stimulated glucose uptake in skeletal muscle in humans: studies using [15O]O2, [15O]H2O, [18F]fluoro-deoxy-glucose, and positron emission tomography. Diabetes 2000; 49:1084-1091. doi: 10.2337/diabetes.49.7.1084.
Phelps ME, Huang S-C, Hoffman EJ, Selin C, Sokoloff L, Kuhl DE. Tomographic measurement of local cerebral glucose metabolic rate in humans with [F-18]2-fluoro-2-deoxy-D-glucose: validation of method. Ann Neurol. 1979; 6: 371-388. doi: 10.1002/ana.410060502.
Phelps ME, Huang S-C, Mazziotta JC, Hawkins RA. Alternate approach for examining stability of the deoxyglucose model lumped constant. J Cereb Blood Flow Metab. 1983; 3(Suppl 1): S13-S14.
Raitakari M, Nuutila P, Ruotsalainen U, Laine H, Teräs M, Iida H, Mäkimattila S, Utriainen T, Oikonen V, Sipilä H, Haaparanta M, Solin O, Wegelius U, Knuuti J, Yki-Järvinen H. Evidence for dissociation of insulin stimulation of blood flow and glucose uptake in human skeletal muscle. Studies using [15O]H2O, [18F]fluoro-2-deoxy-D-glucose, and positron emission tomography. Diabetes 1996; 45: 1471-1477. doi: 10.2337/diab.45.11.1471.
Reivich M, Kuhl D, Wolf A, Greenberg J, Phelps M, Ido T, Casella V, Fowler J, Hoffman E, Alavi A, Som P, Sokoloff L. The [18F]fluorodeoxyglucose method for the measurement of local cerebral glucose utilization in man. Circ Res. 1979; 44: 127-137. PMID: 363301.
Rokka J, Grönroos TJ, Viljanen T, Solin O, Haaparanta-Solin M. HPLC and TLC methods for analysis of [18F]FDG and its metabolites from biological samples. J Chromatogr B 2017; 1048: 140-149. doi: 10.1016/j.jchromb.2017.01.042.
Sasaki H, Kanno I, Murakami M, Shishido F, Uemura K. Tomographic mapping of kinetic rate constants in the fluorodeoxyglucose model using dynamic positron emission tomography. J Cereb Blood Flow Metab. 1986; 6: 447-454. doi: 10.1038/jcbfm.1986.78.
Shankar LK, Hoffman JM, Bacharach S, Graham MM, Karp J, Lammertsma AA, Larson S, Mankoff DA, Siegel BA, Van den Abbeele A, Yap J, Sullivan D; National Cancer Institute. Consensus recommendations for the use of 18F-FDG PET as an indicator of therapeutic response in patients in National Cancer Institute Trials. J Nucl Med. 2006; 47(6): 1059-1066. PMID: 16741317.
Schmidt K, Lucignani G, Moresco RM, Rizzo G, Gilardi MC, Messa C, Colombo F, Fazio F, Sokoloff L. Errors introduced by tissue heterogeneity in estimation of local cerebral glucose utilization with current kinetic models of the [18F]fluorodeoxyglucose method. J Cereb Blood Flow Metab. 1992; 12: 823-834. doi: 10.1038/jcbfm.1992.114.
Shreve PD, Anzai Y, Wahl RL. Pitfalls in oncologic diagnosis with FDG PET imaging: physiologic and benign variants. Radiographics 1999; 19(1): 61-77. doi: 10.1148/radiographics.19.1.g99ja0761.
Slimani L, Oikonen V, Hällsten K, Savisto N, Knuuti J, Nuutila P, Iozzo P. Exercise restores skeletal muscle glucose delivery but not insulin-mediated glucose transport and phosphorylation in obese subjects. J Clin Endocrinol Metab 2006; 91(9): 3394-3403. doi: 10.1210/jc.2006-0269.
Sokoloff L, Reivich M, Kennedy C, Des Rosiers MH, Patlak CS, Pettigrew KD, Sakurada O, Shinohara M. The [14C]deoxyglucose method for the measurement of local cerebral glucose utilization: theory, procedure, and normal values in the conscious and anesthetized albino rat. J Neurochem. 1977; 28(5): 897-916. doi: 10.1111/j.1471-4159.1977.tb10649.x.
Villien M, Wey H-Y, Mandeville JB, Catana C, Polimeni JR, Sander CY, Zürcher NR, Fowler JS, Rosen BR, Hooker JM. Dynamic functional imaging of brain glucose utilization using fPET-FDG. NeuroImage 2014; 100: 192-199. doi: 10.1016/j.neuroimage.2014.06.025.
Vriens D, Visser EP, de Geus-Oei L-F, Oyen WJG. Methodological considerations in quantification of oncological FDG PET studies. Eur J Nucl Med Mol Imaging 2010; 37: 1408-1425. doi: 10.1007/s00259-009-1306-7.
Webb RL, Landau E, Klein D, DiPoce J, Volkin D, Belman J, Voutsinas N, Brenner A. Effects of varying serum glucose levels on 18F-FDG biodistribution. Nucl Med Commun. 2015; 36(7): 717-721. doi: 10.1097/MNM.0000000000000319.
Zhuang H, Codreanu I. Growing applications of FDG PET-CT imaging in non-oncologic conditions. J Biomed Res. 2015; 29(3): 189-202. doi: 10.7555/JBR.29.20140081.
Tags: FDG, Glucose, Lumped constant, Brain, Liver, Skeletal muscle, Bone, Lungs
Updated at: 2023-09-15
Created at: 2009-01-09
Written by: Vesa Oikonen