Determination of glucose consumption Deoxyglucose method

Transcription

1 Determination of glucose consumption Deoxyglucose method Lisbeth Marner, MD, DMSc, PhD Department of Clinical Physiology, Nuclear Medicine and PET Copenhagen University Hospital Rigshospitalet Denmark

2 Glucose metabolism Overview 1. Global measurements of brain glucose metabolism 2. The Deoxyglucose method (DG and FDG) 3. PET-FDG 4. Basal clinical applications of PET-FDG

3 Why study human cerebral glucose metabolism? Glucose is the main energy supply to the brain Brain weighs 2% of body weight, but consumes 20% of the total energy Cerebral Metabolic Rate of glucose (CMRglc)= 24 mmol/min/100g (80g/day) Very stable: Maximal activation leads to a global increase of only 10-15% Only during deep sleep do we see significant reduction (25%) in normal subjects BUT: Regional changes and changes in disease are very useful for our understanding of brain function 80g sugar

4 The coupling of neural activity and energy supply "..an intrinsic mechanism by which its vascular supply can be varied locally in correspondence with variations of functional activity Sir Charles Sherrinton ( ) Seymour S. Kety ( ) "Now, for the first time, the clinical physiologist is no longer at a disadvantage in studying the circulation in the human brain" Charles Schmidt on the Kety-Schmidt method, 1948

5 Brain glucose metabolism clinical aspects 1. Studies of normal brain function Close coupling between neuronal activity and glucose uptake 2. Studies of abnormal brain function Degenerative, cerebrovascular disorders and epilepsy lead to regional reduced glucose uptake

6 Glucose metabolism Glycolysis requires no oxygen Only little energy production Oxygen consuming breakdown of glucose through Krebs cycle and electron chain reactions leads to high energy production Pyruvate can be transformed into lactate, which keeps NADH levels stable

7 Glucose measurements in vivo 1. Global measurements Using global blood flow measurements and Fick s principle 2. Regional measurements using imaging Deoxyglucose method in animals Fluoro-deoxyglucose method in humans

8 THE FICK PRINCIPLE Everything that goes in and doesn t come out again has been taken up by the organ Uptake = F(C a C v ) F= cerebral blood flow C a and C v = substrate concentrations in arterial and cerebral venous blood F can be measured by the Fick Principle by using an inert gas (Xenon)

9 Global measurement of cerebral blood flow by Kety-Schmidt

10 Global measurement of cerebral blood flow by Kety-Schmidt

11 Kety-Schmidt measurement of global blood flow measuring xenon in arterial and venous blood after saturation of brain with xenon Xe- infusion VENOUS ARTERIAL F: Height divided by integrated V-A difference F Olesen, Paulson and Lassen Stroke 2: Q ( C ) V CA

12 Disadvantages: Very invasive! Catheter in the internal jugular vein is necessary to measure cerebral venous blood

13 Catheter in the internal jugular vein

14 Advantages: When CBF is known, almost everything can be measured Ex: Brain Carbohydrate Metabolism after 3.5 Days of Starvation 0,35 0,3 0,25 0,2 0,15 0,1 0,05 p<0,05 Net Uptake (umol/g/min) p<0,05 (n= 9) Control Starvation 0-0,05-0,1 GLU b-ohb AcAc PYR LAC Hasselbalch et al., JCBF, 1994

15 In summery: Brain metabolism calculated by the FICK PRINCIPLE Advantages 1. Reliable, reproducible method 2. Almost any substance can be measured - as long as it is measurable in blood Disadvantages 1. Very invasive 2. No regional information

16 The next major step forward: The deoxyglucose method

17 The deoxyglucose method D-glucose [ 14 C]2-deoxy-D-glucose [ 18 F]2-deoxy-D-glucose = [ 18 F]FDG

18 Blood BBB Brain Glu (Ca) K1 k2 Glu (Ce) hexokinase k3 Glu-6-P (Cm) CO2 + H2O DG (Ca*) K1* k2* DG (Ce*) hexokinase k3* DG-6-P (Cm*) Deoxyglucose is not a substrate for further glycolysis and thus remains trapped in the tissue

19 Compartment model of Deoxy-Glucose metabolism Blood Brain (not metabolized) Brain (metabolized) K1* k3* DG k2* DG (k4*) DG-6-P

20 The autoradiographic method - the Sokoloff equation

21 CMRglc measured by 1 time point = the autoradiographic method Radioactivity measured by scanner and in blood Uptake in brain Total input to brain Time

22 The autoradiographic method - the Sokoloff equation subtracted by small correction terms.. and corrected for difference between tracer and glucose itself Sokoloff et al., 1977

23 % of peak activity in plasma Inherent problems with the deoxyglucose method We need estimation of rate constants (K 1 -k 3 ) to calculate correction terms (luckily they become smaller the longer we wait) Solutions: 1. use norm values from other experiments 2. estimate them from your own experiment 0,160 0,140 0,120 0,100 0,080 0,060 0,040 With full uptake and input curves, we can estimate constants from fitting of model 0,020 0, Time (min)

24 % of peak activity in plasma FDG uptake determined from coefficients (K 1 *-k 3 *) - "the dynamic method" 0,200 0,180 0,160 FDG Clearance: K i * K1 k k k * 2 * 3 * 3 0,140 0,120 0,100 0,080 0,060 0,040 0,020 Flux = K i x C P (mmol/g/min) CMRglc C LC P K k * 1 * 2 k k * 3 * 3 0, Time (min)

25 Lumped Constant (LC) The correction from FDG clearance to glucose clearance can also be calculated from time activity curves LC = Net Clearance of FDG / Net Clearance of Glucose LC K K * i i * K1 k * k2 k K1 k3 k k 2 * 3 * 3 3 K i = net clearance (ml/g/min) LC for FDG in humans: depeding on model Normal litterature values can be used for most studies However, LC changes in hypoglycemia!

26 Lumped Constant (LC)

27 Linearization methods are very often used to calculate CMRglc from PET-FDG 1. Based on compartment models with irreversible binding 2. Clearance (the amount of accumulated tracer in relation to the amount of tracer that has been available in plasma) is measured at equilibrium as the slope of the plot

28 Patlak plot

29 Patlak plot The solution to a two-tissue compartment model (k 4 =0) is: This was rearranged by Patlak and Gjedde: Which after dividing by C P is a straight line when t=t*: P 3 ) ( T 3 2 C k e k k k K C t k k P 0 P i ND P T C d C K V C C t t T d C K C V C 0 P i P ND

30 Patlak plot C C T P V ND K i t 0 C P C P d From the fitted line we therefore have: The metabolic rate K i The distribution volume K1k3 k k 2 V ND 3 is the slope K k k k is the intercept

31 Siemens, Knoxville, USA Parameter maps Patlak Plot SUM SLOPE INTERCEPT

32 Patlak plot K i varies with segment used for determining the slope - why? k 4 * > zero = tracer escapes from the brain, not true irreversible binding

33 Quantitation of rcmrglc When absolute levels are required, PET-FDG is the ONLY quantitative method for rcmrglc PET-FDG is increasingly used non-quantitatively Standard Uptake Values (SUV) Internal Ratios "Hot Spots" V.S. 33% reduction in CMRglc during ketone infusion global changes Temporal rcmrglc reduction in frontotemporal dementia (semantic dementia - regional changes

34 Viability Assessment with FDG PET Metabolism in hypoperfused myocardium Flow (NH 3 ) No metabolism in hypoperfused myocardium Glucose Metabolism (FDG)

35 Basic clinical applications of FDG Oncology whole body PET/CT FDG primary use has replaced many different investigations (i.e., CT of lungs and abdomen, ultrasound) glucose metabolism (upregulated glucose transporter) in most tumor types is several fold higher than in normal tissue

36 The Warburg Effect In oncology the Warburg effect is: Cancer cells can change their metabolism to produce energy from a high glycolysis rate with subsequent lactate acid fermentation in the cytosol rather than by a comparatively low rate of glycolysis followed by oxidation of pyruvate in mitochondria as in most normal cells Malignant fast growing tumor cells typically have glycolytic rates up to 200 times higher even at normal oxygen levels Otto H. Warburg Nobel prize in physiology Int J Biol Sci 2015; 11(12):

37 Correlated to index of proliferation Cochet et al (2012) J Nucl Med 53:512 20

38 SUV AND PROGNOSIS..SUV for prediction of survival in patients with metastatic colorectal cancer. Survival significantly higher among patients with low SUV.. De Geus-Oei et al. Annals of Oncology 2006;

39 SUV AND RESPONSE TO THERAPY Reduction in metabolic activity (SUV) after ONE cycle of chemotherapy correlates with final outcome. Weber et al. J Clin Oncol 2003; 21: MacManus et al. J Clin Oncol 2003; 21: De Geus-Oei et al. J Nucl Med 2007; 48:1592-8

40 Blood glucose level High blood glucose levels due to non-fasting state or diabetes mellitus interfere with FDG uptake in malignant lesions When serum glucose > 8mM SUV in tumor drops from 5.1 to 2.8, p<0.02 While SUV in skeletal muscles increase Patlak-based K i decrease markedly (25%) with higher serum glucose Infusing insulin increase the translocation of GLUT 4 thereby rapidly shunting FDG to organs with a high density of transporters (skeletal and cardiac muscles) Metformin strongly increase the SUV of the small and large intestines due to an increased glucose excretion in the gut

41 Tissue activity curves

42 Tissue activity curves

43 PET-FDG in clinical neuroscience C AD FTD AD PET-FDG is a major tool used in diagnosis and basic research of neurodegenerative diseases Jagust et al., 2007

44 CASE 1: 63 year old woman with depressions. One year with kognitive problems: memory and word-finding difficulties. Her son did not notice memory problems but find her more sentimental. Obj.: Depressed mood. No larger memory troubles or language problems. MMSE: 30/30. Top: FDG-Activity distribution Bottom: Surface projections relative to a normal database

45 CASE 2: 58 year old well educated woman with one year of increasing headache, fatique and memory problems. Her daughter thinks that problems started two years ago. Obj.: Learning and memory disabilities, reduced pace and visuoconstructional deficits. MMSE: 28/30. Top: FDG-Activity distribution Bottom: Surface projections relative to a normal database

46 CASE 3: 66 year old woman with depression 3 years earlier now present with slow movements and rigidity. Spouse find increasing symptoms after recuction in antidepressant medicine. Obj.: Obj.: Chronically fatiqued, depressed mood and memory problems. Bradykinesia and a slight tremor. MMSE: 22/30. Top: FDG-Activity distribution Bottom: Surface projections relative to a normal database

47 CASE 4: 73 year old well educated woman. Several years of increasing problems with concentration, language and memeory. Anti-depressant treatment due to anxiety. Obj.: Well maintained and insight to her problems. Word finding difficulties and problems with structure. MMSE: 19/30. Top: FDG-Activity distribution and Surface projections relative to a normal database Bottom: MRI

48 Epilepsia Male: 18 year, debut: 9 years Treatment resistant seizures EEG: Focus temporo-centro-parietal MRI: normal Seizures Tingling sensation in the left side of the mouth Head-turn to the left Left arm lifting Fall backwards Ian Law, PET & Cyclotron Unit, Rigshospitalet

49

50

51 Summery: PET-FDG Well validated kinetic model Allows for quantification of regional CMRglc Invasive (arterial input) Look out for errors (K*'s and LC) Today is primarily used non-quantitatively dementia oncology Great potential in basal physiological neuroscience Tight CMRglc-CBF coupling renders fmri more applicable for functional activation studies