<<

Publications · Brochures

Brain Imaging

A Technologist’s Guide

Produced with the kind Support of Editors

Fragoso Costa, Pedro (Oldenburg) Santos, Andrea (Lisbon) Vidovič, Borut (Munich)

Contributors

Arbizu Lostao, Javier Pagani, Marco Barthel, Henryk Payoux, Pierre Boehm, Torsten Pepe, Giovanna Calapaquí-Terán, Adriana Peștean, Claudiu Delgado-Bolton, Roberto Sabri, Osama Garibotto, Valentina Sočan, Aljaž Grmek, Marko Sousa, Eva Hackett, Elizabeth Testanera, Giorgio Hoffmann, Karl Titus Tiepolt, Solveig Law, Ian van de Giessen, Elsmarieke Lucena, Filipa Vaz, Tânia Morbelli, Silvia Werner, Peter (*) Article was written with the kind support support was written with thekind (*) Article 126 Imprint Chapter 10 Chapter 9 Chapter 8 Chapter 7 Chapter 6 Chapter 5 Chapter 4 Chapter 3 Chapter 2 Chapter 1 5 Introduction Foreword 4 Contents of andincooperation with: Claudiu Peștean Health Care inPatients withNeurological Disorders Grmek Marko Death Karl T. Barthel, Henryk HoffmannandOsamaSabri Solveig Tiepolt, TorstenBoehm, Peter Werner, forPET/MRI Brain Imaging Calapaquí-Terán Adriana K. Delgado-Bolton, Roberto andJavier Arbizu inRadiotherapy Planning ofBrain PET/CT Tumours Ian Law, Valentina andMarco Pagani Garibotto inNeurologicalImaging and Vascular Brain (PET/CT) Diseases Filipa Lucena, and Eva Sousa Tânia F. Vaz inNeurologicalImaging and Vascular andSPECT/CT) Brain (SPECT Diseases Giorgio Testanera andGiovanna Pepe inOncologicalImaging Brain PET/CT Diseases: Elizabeth C.Hackett inOncological Brain (*) Imaging andSPECT/CT SPECT Aljaz Socan Tracers for Brain Imaging andPierre vandeGiessen,SilviaMorbelli PayouxElsmarieke Anatomy, Physiology andPathology Andrea Santos, Pedro Fragoso Costa

3 116 110 100 92 72 54 33 26 12 6 Printed in accordance with the Austrian Eco-Label for printed matters. EANM Foreword

During the past two decades the European becoming standards as diagnostic tools. Association of (EANM) has Following its recent development, PET/MRI become a leading institution for research has also found many suitable clinical ap- and education in all aspects of nuclear med- plications in brain imaging that have helped icine (NM), including tracer development, this new hybrid technique to find a role in scanning technology and multimodality and clinical practice and research protocols. hybrid imaging. Within the last 5 years, the EANM has followed the “Go Clinical” motto, I would like to thank all those who have with the intention of advancing the role contributed to this project as authors and of NM into clinical and research protocols. reviewers, without whom the book would This challenge was accepted by the EANM not have been possible. I would also like to Technologist Committee (TC), which has thank SNMMITS (Society of Nuclear Medi- attempted to ensure that technologists’ cine and Molecular Imaging Technologist education, research and daily practice have Section) and the EANM a strong clinical impact. This intention is also Committee for their help and high-quality reflected in the decision to dedicate the contributions. Technologist Guide 2015 to brain imaging. Special thanks are due to Pedro Fragoso EANM Technologist guides are developed Costa, Andrea Santos and Borut Vidovic for and printed on a yearly basis for presenta- their incredible enthusiasm and compe- tion at the EANM Annual Congress. They tence in dealing with the editorial duties are specifically aimed at NM technologists, and organisational work. Finally I remain ex- radiographers and all other healthcare pro- tremely grateful to the EANM Board, EANM fessionals working or intending to work in a Executive Office, Technologist Committee nuclear medicine department. and all the EANM committees involved in the project. After dedicating three books specifically to PET/CT, we have focussed on specific clini- With my warmest regards, cal fields in which NM has a strong impact on clinical procedures, such as myocardial Giorgio Testanera perfusion imaging and brain imaging. Brain Chair, EANM Technologist Committee imaging was strongly influenced by the PET/CT revolution and is now also a field of high interest for research, thanks to the development of 11C tracers, somatostatin analogues and amyloid-based tracers, which are rapidly

4 the demandingnature oftheexaminations. neurological disorders to cooperate andalso ofpatients sufferinging to from theinability challenging ow- care, whichcanbeparticularly technologists role alsoplay inpatient akey protocol theneuroimaging design. In context, piece ofequipmentrequires optimisationof nologists, given technique that each new and These advancesrepresent achallengefor tech- movement disorders andbraintumours. of conditions, including ,epilepsy, hasbeenstudiedinavariety brain function . From metabolismto perfusion, come increasingly acute and to cover more imaging to be and PET has allowed SPECT over, development radiopharmaceutical ofthesetechniques.clinical impact More the imagesacquired, thereby enhancingthe developed in order to improve of the quality technologiesnew and techniques have been clinical tool inneurology. recent During years, system to anestablished the centralnervous anddisease of offunction in theevaluation have evolved from beingaresearch modality Diagnostic nuclear medicine investigations conditions, includingneurological disorders [1]. chronic andprogressive physical andmental continued increases intheprevalence ofmany to willcontribute and developing countries of thegeneralpopulationinbothdeveloped greater andthe overall ageing life expectancy er’s diseaseorotherdementias. Furthermore, lepsy andaround 35.6 whom more than50 hundreds ofmillionspeopleworldwide, of Neurological disorders are already affecting Andrea Santos, Pedro Fragoso Costa Introduction million suffer from epi- million have Alzheim- - - 5 Reference: Pedro Fragoso Costa Andrea Santos been fundamentalto thecreation of thisbook. offered whichhave theirtimeandexpertise, to thankalltheauthorswhohavelike kindly The EANM Technologist Committee would field. in thisparticular physicians, physicists andstudentsinterested staff, includingnotonly technologists butalso willprovide avaluabletool toof art allclinical roimaging techniques andtheclinical state cal disorders. ofneu- overview This extended lance needsofpatientsaffected by neurologi- devoted- to thespecialhealthcare andsurveil inChapter is described aging inthecaseofsuspicionbraindeath imaging isapproached inChapter 8.Brainim- the emerging for brain technology ofPET/MRI is discussed in Chapter 7. The application of inbraintumourradiotherapy planning PET/CT (Chapters then onPET/CT technology and andSPECT/CT on SPECT brain diseasesisalsoanalysed, focussing first inneurologicaland 4).Imaging andvascular techniques (Chapters and PET/CT SPECT/CT or of oncological diseaseby meansofSPECT chapters ontheimaging provide anoverview are discussed(Chapter (Chapter brain’s anatomy, physiology andpathology we begin the by thisbook, describing In 1. Geneva: Geneva: World HealthOrganisation, 2006. Neurological disorders: Public healthchallenges. 1). After this, tracersfor brainimaging 2). The following two 9. The finalchapter is 5 and 6). The useof 3 3 EANM Chapter 1 Anatomy, Physiology and Elsmarieke van de Giessen, Silvia Morbelli and Pierre Payoux

The brain serves many important functions, largest structure of the brain and is also divid- including movements, thoughts, learning, ed into left and right hemispheres, which are memory, language, integration and inter- connected through the corpus callosum. The pretation of our senses, and critical functions cerebrum is covered with the cerebral cortex such as respiration. To understand imaging that is folded into gyri and sulci to increase its of the brain it is essential to have knowledge surface. The cerebral cortex can be divided of the anatomy and its normal function. This into four lobes: frontal, parietal, temporal and chapter provides an overview and describes occipital. The frontal lobes are located in the several common brain pathologies. front of the brain and include the primary mo- tor cortex and , which is im- Anatomy portant for cognitive and executive functions. Brain structures The temporal lobes are located on both sides The brain consists of three major structures: the of the brain and contain the auditory cortex. brainstem, and cerebrum (Fig. 1). On the medial sides of the temporal lobes are The brainstem is the lower structure of the the , important for memory for- brain, extending into the spinal cord at the mation, and the , involved in emo- point where it leaves the skull. It is an impor- tions. The occipital lobes are at the back of the tant relay station and consists of the , brain and are essential for processing visual pons and medulla oblongata. The medulla information. The parietal lobes are located oblongata is important for primitive functions in between the frontal and occipital lobes; like respiration, blood pressure and rate. they include the somatosensory cortex and are involved in integrating sensory informa- tion to form a perception. The are located subcortically, close to the centre of the cerebrum. They play an important role in controlling and regulating activities of the motor system and in motivation and habitu- ation. A major structure of the basal ganglia is the , which is involved in the plan- ning and modulation of movement pathways, (figure in public domain) (figure but is also part of the cognitive processes in- Figure 1: The human brain volving executive function, such as working memory. The is located superior to The cerebellum is located at the back of the the midbrain, near the centre of the brain. This brain and is divided into two hemispheres is an important relay station for signals to the (left and right). It is important for coordination cerebral cortex, including motor and sensory and timing of movement. The cerebrum is the signals, but it is also involved in sleep/wakeful-

6 Chapter 1 Anatomy, Physiology and Pathology

ness and consciousness. Below the thalamus posterior cerebral arteries. The venous drainage lies the hypothalamus, which is connected to consists of the venous sinuses in the dura mater the pituitary gland. The hypothalamus is con- and of veins inside the deep structures of the sidered to be the link between the nervous brain, which join behind the midbrain and then system and the endocrine system, regulating drain into the sinus system. The sinus drainage appetite, electrolyte balance and body tem- connects to the jugular veins in the neck. perature. The two major components of the brain structures are grey and white matter. Physiology Grey matter largely consists of neuronal cell Cell types bodies and white matter is white because of Two main types of cell compose the brain: the large number of myelinated axons. neurons and glial cells. Neurons transmit nerve signals to and from the brain over

Meninges and cerebrospinal fluid long distances. They consist of a cell body EANM The brain and spinal cord are enveloped and (or soma) with branching dendrites (signal protected by the meninges, which consist receivers) and a projection called an axon, of three layers: dura mater, arachnoid mater which conducts the nerve signal and trans- and pia mater. The space in between the pia mits it across a synapse (Fig. 2). mater and the arachnoid mater is termed the subarachnoid space, which is filled with ce- rebrospinal fluid (CSF). CSF is produced from arterial blood by the choroid plexuses. There it is filtered through the blood-brain barrier, which protects the brain from and toxins. The CSF flows through the ventricu- lar system, consisting of the lateral ventricles and third and fourth ventricles.

Cerebral vasculature Two pairs of arteries provide the arterial blood supply of the brain: the internal carotid arter- ies in the neck and the vertebral arteries. They are interconnected through the circle of Wil- lis, which ensures a back-up circulation in the event of dysfunction of one of the arteries. The in public domain) (figure three main pairs of arteries that branch off from Figure 2: Propagation of a signal down an the circle of Willis and supply the cerebrum axon to the cell body and dendrites of the with arterial blood are the anterior, middle and next cell

7 Neurons can be classified according to the the sympathetic nervous system. The nor- number of extensions deriving from the adrenergic system effects are alertness and neuron’s cell body (bipolar, pseudo-unipolar, arousal and influences on the reward system. multipolar) or according to the direction in is an inhibitory neurotransmit- which they send information (afferent/sen- ter involved in the control of locomotion, sory, efferent/motor and interneuron). cognition, affect and neuroendocrine secre- tion. Dopamine deficiency results in Parkin- Glial cells have supportive functions in that son’s disease. GABA (γ-aminobutyric acid) is they help to define synaptic contacts and to the main inhibitory neurotransmitter in the maintain the signalling abilities of neurons, human . The GAB- but they do not participate directly in elec- Aergic system is the target of many drugs trical signalling. There are three types of glial and substances including , cell in the mature central nervous system: and . Glutamate is the astrocytes, oligodendrocytes and microglial most common neurotransmitter in the CNS cells. and is involved in most aspects of normal brain function, including cognition, mem- , neurotransmitters and ory and learning. is an inhibitory transporters neurotransmitter that has been found to be The brain relies entirely on for its en- intimately involved in emotion, mood, ap- ergy supply and uses approximately 20% of petite and sleep. Modulation of serotonin at the body’s total energy requirements. Mea- synapses is thought to be a major action of surements of glucose consumption reflect several classes of pharmacological antide- the amount of brain activity in the various re- pressant. gions of the brain and allow us to learn more about how the brain works. Neurotransmitter transporters are a class of membrane transport proteins that span the Neurotransmitters are the chemicals which cellular membranes of neurons. A variety of allow the transmission of signals from one neurotransmitter re-uptake transporters are neuron to the next across the synaptic cleft. pharmacotherapeutic targets for modulating is responsible for muscle the synaptic neurotransmitter concentration, stimulation and has been found in the auto- and therefore neurotransmission. Among nomic nervous system and sensory neurons. neurotransmitter transporters, the dopamine The cholinergic system modulates cognitive transporter (DAT) has a crucial role for nucle- performance and gait stability and has been ar imaging of the system. DAT linked with cognitive/motor dysfunction in pumps dopamine out of the synapse back several neurodegenerative diseases. Nor- into cytosol, from which other transporters epinephrine or noradrenaline is prevalent in sequester dopamine and noradrenaline into

8 Chapter 1 Anatomy, Physiology and Pathology

vesicles for later storage and release. are central features. Fluctuating cognition, visual hallucination and spontaneous fea- Pathology tures of parkinsonism are typical. Finally, REM Neurodegenerative disease sleep behaviour disorder, severe sensitiv- Alzheimer’s disease (AD) is the most com- ity to neuroleptics and low dopamine trans- mon cause of dementia among older people. porter uptake in the basal ganglia as seen Dementia is the loss of memory and other on SPECT/PET imaging scans are suggestive intellectual functions of sufficient sever- features [4]. ity to cause problems in patients’ abilities to perform their usual personal, social or occu- Frontotemporal degeneration (FTD) refers to pational activities. AD is pathologically char- a group of diseases involving the temporal acterised by the presence of plaques com- and/or frontal lobes. FTD is also commonly posed of amyloid-beta peptides and neuro- referred to as frontotemporal dementia and EANM fibrillary tangles containing phosphorylated occurs most frequently in persons under tau [1]. Enzymatic cleavage of amyloid-beta the age of 65. FTD has a heterogeneous peptides is considered to play an important spectrum, with behavioural, cognitive or lan- pathogenetic role in AD (likely occurring sev- guage changes [5]. eral years before clinical manifestations) [1]. The ongoing challenge is the distinction of Parkinson’s disease (PD) is an idiopathic de- slight cognitive deficit which precedes AD generative disease which results from the dementia from normal aging. Mild cognitive death of dopamine-generating cells in the impairment (MCI) refers to a subtle but mea- substantia nigra. PD usually affects people surable cognitive decline in the absence of over the age of 50. Symptoms of PD are rest- dementia and those affected include both ing tremor, rigidity, bradykinesia and postural subjects who remain stable over time and instability. Onset of symptoms is gradual and patients who become demented due to typically asymmetrical [6]. underlying neurodegenerative aetiologies, including AD [2]. Imaging and CSF biomark- Cerebrovascular disease ers allow the identification of probable AD Cerebrovascular disease (CVD) refers to a among MCI patients with a typical gradual group of conditions that impair the circula- and progressive (isolated or not) episodic tion of blood to the brain, causing limited memory impairment [3]. or no blood flow to affected areas. Clinical symptoms depend on the location and ex- Dementia with Lewy bodies (DLB) is a degen- tent of cerebral tissue affected. CVDs are erative dementia of unknown aetiology. It is subdivided into ischaemic events and ce- the second most common form of dementia. rebral haemorrhages. Ischaemic events are Deficits in attention and executive function further classified according to the duration

9 of symptoms (transient ischaemic attacks Epilepsy last more than a few minutes and less than Epilepsy is characterised by sudden recurrent 24 h; ischaemic stroke lasts more than 24 h episodes of sensory disturbance, loss of con- and may be progressive stable, or resolving. sciousness or convulsions, associated with In some cases a number of small strokes or abnormal electrical activity in the brain caus- a large stroke may cause so-called vascular ing (a transient occurrence of signs dementia (the second most common form and symptoms due to abnormal excessive or of dementia after AD) [7]. synchronous neuronal activity) [10]. can be focal or generalised. The cause of Brain tumours most cases of epilepsy is unknown (“crypto- Primary brain tumours are classified accord- genic”) although some people develop epi- ing to the type of tissue in which they arise lepsy as the result of brain injury, stroke, brain [8]. The most common brain tumours are tumour or substance abuse. gliomas, which begin in the glial (supportive) tissue. There are several types of glioma, in- cluding astrocytomas, oligodendrogliomas and ependymomas. Other types of brain tumour that do not begin in glial tissue are meningiomas, schwannomas, craniopha- ryngiomas, germ cell tumours and pineal region tumours. Cerebral secondary tumours are metastatic tumours originating in other organs and are more common than primary ones (8).

Brain and infections Encephalitis refers to a diffuse brain paren- chymal inflammation mainly due to viral infections. Meningitis is an inflammation of meninges surrounding both brain and spinal cord. It can be caused by viral, bacte- rial and fungal infections. Symptoms include headache, fever, confusion, drowsiness and fatigue, and in some cases seizures or con- vulsions, hallucinations, stroke, haemorrhag- ing and memory problems occur [9].

10 1. References References Chapter 1 3. 2. 10. 9. 8. 7. 6. 5. 4. Dubois B, Feldman HH,Jacova C,HampelH,Molinuevo andoutcome.terization Arch Neurol. 1999:56:303–8. cognitive clinical charac- impairment: E. Mild Kokmen Petersen SmithGE, RC, Waring RJ, SC,Ivnik Tangalos EG, 1998;54:77–95. er’s update. disease:acritical JNeural Transm Suppl. Bancher C. Neuropathology KA, of Alzheim Jellinger 2003;349:1257–66. Chang BS,Lowenstein DH.Epilepsy. NEnglJMed. 2009;73:e14–5. neuroinfectious diseases.in neurology: Neurology. JJ, EpsteinMillichap LG. Emerging subspecialties 2011;84:S90–106. J Radiol. Biology, genetics and imaging of glial cell tumours. Br Walker A,Crooks C,Baborie D, MD. S,Jenkinson Wilkins DSM-IV, NINDS-AIREN). Stroke. 1996;27:30–6. ICD-10, ent diagnostic for criteria vasculardementia(ADDTC, Wetterling T, RD, ofdiffer Comparison Kanitz Borgis KJ. Disord.Relat 2012;18Suppl1:S31–3. Brooks DJ. Parkinson’s disease: diagnosis. Parkinsonism dementia. Brain.2011;134:2456–77. forcriteria offrontotemporal thebehavioural variant er JH,NeuhausJ, ofrevised etal. diagnostic Sensitivity D, HodgesJR,Knopman MF, Mendez K, Rascovsky Kram Neurology.DLB Consortium. 2005;65:1863–72. ment ofdementia with Lewy of the bodies: third report onDLB.man H,etal.; Diagnosis andmanage Consortium IG,DicksonDW,McKeith Lowe J, JT, Emre M,O’Brien Feld Neurol. 2014;13:614–29. forteria Alzheimer’s Lancet criteria. disease:theIWG-2 etal. AdvancingJL, Blennow research K, diagnostic- cri - - - - - 11

EANM Chapter 2 Tracers for Brain Imaging Aljaz Socan

Introduction diffusion or by specific transport mecha- emission tomography (PET) and nisms. Depending on the brain target of the single-photon emission computed tomogra- imaging and the attendant required tracer phy (SPECT) imaging use positron- and gam- properties, tracers for brain imaging can be ma-emitting radioisotopes that can easily be divided into several groups: regional cerebral incorporated into biological molecules and blood flow (rCBF) tracers, metabolic tracers, thus allow the measurement of functional tracers targeting neurotransmission and re- parameters (physiological and/or pharma- ceptors, tracers targeting amyloid and tracers cological interactions) of tissue rather than for brain tumour imaging. just providing the anatomical definition of structures. Both techniques are exceptionally rCBF tracers sensitive (PET more so than SPECT); they can In order to be used for the measurement detect picomolar or even femtomolar con- of brain perfusion, the radiopharmaceutical centrations of radiolabelled compounds and must possess certain physiological proper- enable the dynamic acquisition of relatively ties. In addition to being able to cross the fast kinetics (of the order of seconds for PET). BBB, their extraction must approximate unity With these properties, PET/SPECT can facili- and must be independent of blood flow; as a tate the quantitative measurement of rapid consequence their initial distribution will be physiological/pharmacological processes of proportional to rCBF. They also must be re- biomolecules in the living brain, for example tained within the brain in their initial distribu- [1]. The increase in neurological applications tion for sufficiently long to enable diagnostic of PET/SPECT over the past decade has been images to be obtained [3]. Ideally, tracer up- greatly aided by the significant improve- take should show no redistribution, so that ments in data acquisition (hardware technol- the initial uptake, reflecting rCBF at a fast ogy) and data quantification (model-based time window after injection, will remain al- methodology) and by the vastly growing most unchanged for several hours. The result number of available PET/SPECT metabolic ra- is images independent of variations in rCBF diotracers and receptor-specific . after the fixation time. Two tracers are cur- Table 1 lists the more common PET/SPECT ra- rently used for evaluation of rCBF in routine diotracers/ligands and their intended targets clinical practice: 99mTc-HMPAO (hexamethyl in clinical neurology [1,2]. propylene amine oxime, exametazime, Ceretec®) and 99mTc-ECD (ethyl cysteine di- For the tracer (radiopharmaceutical) to be mer, bicisate, Neurolite®); these are discussed able to enter the CNS (brain), it is a prereq- further below. Other available tracers such as 123 15 uisite that it possesses the capacity (suit- I-IMP and H2 O are still used mainly for re- able physiochemical properties) to cross the search purposes. blood-brain barrier (BBB), whether by free

12 Chapter 2 Tracers for Brain Imaging

Brain perfusion SPECT using 99mTc-labelled 99mTc-HMPAO primary complex is uncharged, lipophilic and of sufficiently low molecular SPECT is a technique that produces topo- weight to cross the BBB readily. However, it graphic images of the three-dimensional converts, at a rate of approximately 12% per distribution of a radiopharmaceutical. Ap- hour, to a less lipophilic secondary complex plied to the brain, it can be used to mea- which does not cross the BBB. This would sure regional cerebral perfusion [3]. The two limit the useful shelf-life of the product to tracers most widely used in brain SPECT in 30 min but the addition of cobalt stabiliser, the EU, 99mTc-HMPAO and 99mTc-ECD, differ in 1-5 min after preparation of radiopharma- terms of their in vitro stability, uptake mech- ceutical, lengthens its shelf-life to 5 h. The anism, cerebral distribution and dosimetry; adult dose is 350–500 MBq. The primary nevertheless, their kinetic properties are complex clears rapidly from the blood after very similar. They enter the brain cells ow- intravenous injection. Uptake in the brain EANM ing to their lipophilicity and then remain reaches a maximum of 3.5–7% of the in- there as a consequence of their metabolic jected dose within 1 min after injection. Up conversion into hydrophilic compounds. In to 15% of the cerebral activity washes out of the case of ECD, the stable lipophilic com- the brain by 2 min post injection, after which plex crosses the BBB by passive diffusion. there is slight loss of activity for the following Its localisation in the brain depends upon 24 h due to the physical decay of 99mTc. Ac- both the perfusion of the region and the tivity not associated with the brain is distrib- uptake of the tracer by the cells. Once in the uted widely throughout the body, especially brain cells, the complex is metabolised to in the muscle and soft tissue. Approximately polar, less diffusable compounds, complex 30% of the injected dose is found in the GI de-esterification being the crucial reaction tract immediately after injection, with about leading to hydrophilic conversion. After 50% of this being excreted over 48 h. In addi- background , brain images may tion, over a 48-h period, 40% of the injected be obtained from 10 min to 6 h after injec- dose is excreted via the kidneys and urine [1]. tion of the adult dose of 370–1110 MBq, the optimum imaging time being 30–60 min af- The different retention mechanisms of 99mTc- ter injection. The primary route of ECD and 99mTc-HMPAO are the reasons for the is via the kidneys, with 73% of the injected different behaviour of the tracers in specific dose being cleared through the bladder disorders; for example, in subacute stroke during the first 24 h (up to 50% is cleared the distribution of 99mTc-ECD seems to reflect within the first 2 h). Approximately 11% of metabolic activity more closely, while 99mTc- the injected dose is eliminated via the GI HMPAO is better correlated with cerebral tract over 48 h [1]. perfusion (Fig. 1) [1]. The tracers are not inter- changeable, and 99mTc-ECD and 99mTc-HMPAO

13 SPECT do not provide absolute quantitative various regions of the brain. Normally, there blood flow values but rather estimate rela- is intense uptake of tracer in the grey matter tive regional flow differences based on the of the cortex and basal ganglia, with signifi- comparison of count density ratios between cantly lower uptake in the white matter.

Figure 1: Brain perfusion SPECT imaging with 99mTc-HMPAO in a patient with suspected brain death after CNS trauma. There is complete absence of brain perfusion (cerebrum, cerebellum, brain stem)

14 Chapter 2 Tracers for Brain Imaging

Common indications for 99mTc-ECD and 99mTc- is studied with analogues of glucose. 2-(18F) HMPAO according to the EANM guidelines Fluoro-2-deoxy-D-glucose ([18F]FDG) today for brain perfusion SPECT are: represents the workhorse of imaging in sev- eral fields of PET nuclear medicine. 18[ F]FDG • Evaluation of cerebrovascular disease: is a glucose analogue and is therefore taken up in cells via glucose transporters (glucose • Acute stroke does not freely cross the BBB or cell mem- • Chronic ischaemia (chronic brane); it is then phosphorylated by hexoki- cerebrovascular disease) nase into FDG-6-phosphate, which cannot • Preoperative evaluation for potential be metabolised (unlike glucose) and is con- ischaemia following carotid artery sequently trapped in the cell (Fig. 2) [1]. sacrifice EANM • Postsurgical lateralisation and localisation of epileptogenic foci • Evaluation of suspected dementia (early detection and differential diagnosis of various forms of dementia) • Evaluation of traumatic brain injury (adopted from Reports, from (adopted Spandidaos Publications) • Evaluation of suspected inflammation

• Assessment of brain death Figure 2: Mechanism of metabolic trapping of 18 Prior to the investigation, patients should [ F] in cells preferably avoid excessive (caf- Following intravenous administration of feine, cola, energy drinks), alcohol, smoking 100–400 MBq, the pharmacokinetic profile and any other drugs known to affect the CBF. of [18F]FDG in the vascular compartment is (It is necessary to discuss drug withdrawal bi-exponential. It has a distribution half-life with the clinician caring for the patient.) of 1 min, a large volume of distribution and an elimination half-life of approximately Metabolic tracers 12 min without being metabolised. Elimina- The main metabolic substrates of the brain tion is mainly renal. Approximately 20% of are oxygen and glucose. Both metabolic the injected dose is excreted in urine dur- pathways are studied with specific tracers. ing the first 2 h. After administration, FDG Oxygen-15 is the tracer used in research to is distributed mainly to the brain and heart. study the regional cerebral metabolic rate of Approximately 7% of the injected dose is oxygen (rCMRO2), while glucose metabolism

15 accumulated in the brain within 80–100 min Normally, there is high tracer uptake in the after injection. Approximately 3% of the grey matter of the cortex and basal ganglia; injected activity is taken up by the myo- uptake in the white matter is significantly cardium within 40 min, while ca. 0.3% and lower. Common indications for [18F]FDG in 0.9%–2.4% is accumulated in the pancreas PET brain imaging according to the EANM and lungs, respectively [1]. In the brain, glu- procedure guidelines for PET brain imaging cose metabolism provides approximately using [18F]FDG [4] are: 95% of the ATP required for brain function and is tightly connected to neuronal activ- • Dementing disorders: early diagnosis ity both in physiological conditions and in and differential diagnosis of dementing several diseases affecting the brain. Changes disorders, such as Alzheimer’s disease in neuronal activity induced by disease are and frontotemporal dementia. The typical reflected in an alteration in glucose metabo- topographic patterns of hypometabolism lism. Since the accumulation of [18F]FDG in may help to diagnose the main brain tissue depends on facilitated transport neurodegenerative diseases at the of glucose and -mediated phos- predementia stage. phorylation, it is suitable for imaging of re- • Neuro-oncology: differential diagnosis gional cerebral glucose consumption and is of cerebral space-occupying lesions, currently the most accurate in vivo method detection of viable tumour tissue for the investigation of regional human brain (recurrence), non-invasive grading. metabolism. Its clinical use can be regarded as established for a number of diagnostic • Epilepsy (interictal injection): preoperative questions in neurology, neurosurgery and evaluation of partial epilepsy in adults and psychiatry [4]. children to identify the functional deficit zone. Depending on the clinical question and the • Movement disorders: differentiation type of equipment, [18F]FDG imaging may between Parkinson’s disease and atypical include static limited field tomographic im- parkinsonian syndromes. ages or dynamic tomographic images (used when absolute quantification of regional Tracers targeting the dopaminergic system metabolic rates of glucose is needed), 2D or (neurotransmission and receptors) 3D acquisition mode and attenuation cor- The synaptic cleft is the location where in- rection (mandatory for [18F]FDG PET brain teractions between cells take place. The imaging). neurotransmitter released by the presyn- aptic neuron reaches the postsynaptic cell

16 Chapter 2 Tracers for Brain Imaging

membrane, where receptors are present, ogy or disease and under the effect of drugs. and initiates the neurotransmission chain of All three are possible targets of imaging with events. The transmitter can also re-enter the radiolabelled tracers. On account of the im- presynaptic neuron via the reuptake chan- portance of the brain functions connected nels, which actively participate in modu- with its integrity (its dysfunction leads, for lating the intracleft concentration. All the example, to movement disorders and cogni- constituents of the synaptic transmission tive decline), the dopaminergic system is the chain, i.e. transmitter, receptors and reuptake most extensively studied neurotransmitter channels, can be modulated by functional system in brain nuclear medicine imaging. requests as a consequence of local physiol- EANM (adopted from Brain, Oxford Journals) Brain, Oxford from (adopted Figure 3: Dopaminergic radioligands for SPECT and PET

Various SPECT and PET dopaminergic trac- and 99mTc ([123I]β-CIT, [123I]FP-CIT and [99mTc] ers are currently available for routine clinical TRODAT-1) are the most frequently used trac- practice and research (Fig. 3). [18F] ers for the study of nigrostriatal dopaminer- and derivatives labelled with 123I gic dysfunction in Parkinson’s disease (Fig. 4).

17 A B

Figure 4 A, B: Dopamine reuptake transporter (DAT) SPECT imaging ([123I]FP-CIT). (A) Normal uptake in the striatum (, ). (B) Patient with Parkinson’s disease: asym- metrically decreased uptake is seen in the striatum (more pronounced on the right side). There is bilaterally reduced uptake in the putamen posteriorly

PET/SPECT can also be used non-invasively disorders (D1 and D2 receptor ligands) and for to indirectly monitor changes in neurotrans- pharmacological studies on endogenous do- mitter concentration, providing that a PET pamine changes (D2 receptor ligands). specific and selective for the system of interest is available and the radio- 6-[18F]Fluorolevodopa is a metabolic tracer binds to the same site as the endog- that is taken up into dopaminergic nerve ter- 18 enous ligand or neurotransmitter. The D2 re- minals and converted into [ F]dopamine. The ceptor radioligands [11C] and [123I] radiotracer can be used as a measure of do- IBZM are, for example, particularly sensitive pamine synthesis and dopaminergic neuron to changes in dopamine levels. Dopamine density and, therefore, presynaptic dopami- receptor ligands have been used for the nergic function. In Parkinson’s disease, con- identification of postsynaptic dopaminergic ventional MRI is unable to identify any ana- deficit in parkinsonian neurodegenerative tomical abnormalities. With [18F]F-DOPA PET, disorders (D2 receptor ligands), for the as- the loss of dopaminergic neurons is clearly sessment of neuroreceptor/neurochemical detected in the striata and has been shown to changes and drug occupancy in psychiatric be useful in early differentiation of Parkinson’s

18 Chapter 2 Tracers for Brain Imaging

disease from other movement disorders. In for VMAT-2 and 11C-PE21 for DAT. Various addition, [18F]F-DOPA can be used to detect cocaine analogues that are labelled with 123I a subclinical parkinsonian-like pattern (i.e. and are suitable for SPECT have been shown degeneration of dopaminergic neurons) in as- to bind with high affinity to DAT. Currently ymptomatic adult identical twins [1]. β-CIT (DOPAS-CAN®) and FP-CIT (, DaTSCAN®) are commercially available and Brain neurotransmission SPECT using widely used for clinical evaluation of patients 123I-labelled ligands in Europe [5]. The administered dose of [123I] The dopaminergic neurotransmitter system ioflupane in adults is 150–250 MBq. 123[ I]Ioflu- plays a major role in movement disorders, pane is cleared rapidly from the blood, with particularly in parkinsonism and dementia only 5% remaining at 5 min after intravenous with Lewy bodies. The nigrostriatal dopami- injection. Uptake in the brain is rapid, with nergic pathway is best analysed at the striatal 7% of the injected activity being present in EANM level, where the nigrostriatal neurons end the brain after 10 min, decreasing to 3% after and connect to the postsynaptic nerve ter- 5 h. The primary route of excretion is via the minals using dopamine as a neurotransmit- kidneys, with 60% of the injected dose being ter which binds to the postsynaptic D1 and excreted in the urine at 48 h post injection;

D2 receptors. Both pre- and postsynaptic faecal excretion is calculated at approximate- levels can be targeted by PET or SPECT trac- ly 14% [1]. Normally, there is intense uptake ers. Presynaptic events can be summarised in the striatal structures (caudate nucleus as follows. Dopamine is stored in vesicles and putamen). before being released into the synaptic cleft. Vesicular monoamine transporter 2 (VMAT-2) According to the EANM Guidelines [5], com- transports cytosolic and newly synthesised mon indications for brain neurotransmission dopamine into vesicles. The sodium-depen- SPECT using 123I-labelled dopamine trans- dent dopamine active transporter (DAT), porter ligands are: which is located in the membrane of the pre- synaptic nigrostriatal nerve terminals, is re- • Detection of loss of functional dopami- sponsible for the reuptake of dopamine from nergic neuron terminals in the striatum of the synaptic cleft. L-DOPA is taken up by the patients with clinically uncertain parkinso- presynaptic neurons via amino acid trans- nian syndrome; differentiation of essential porter and decarboxylated to dopamine by tremor from Parkinson’s disease, multiple L-aromatic acid decarboxylase (Fig. 3) [5]. system atrophy and progressive supra- nuclear palsy (on its own, it is unable to Various PET tracers have been used to study discriminate between Parkinson’s disease, these presynaptic targets, such as 18F-DOPA multiple system atrophy and supranuclear for aromatic acid decarboxylase, 11C-DTBZ palsy)

19 • Differentiation of dementia with Lewy ceptor, since the majority of D2 receptors are bodies from other located postsynaptically. According to their pharmacological response, dopamine recep- • Early diagnosis of neurodegenerative par- tors are divided into D -like receptors (D , D ) kinsonism; assessment of the presynaptic 1 1 5 and D -like receptors (D , D and D ). Assess- deficit in early Parkinson’s disease 2 2 3 4 ment of D1-like receptors has not gained any • Assessment of disease severity (DAT bind- clinical significance, but a number of clinical ing is related to the clinical stage and investigations have focussed on the D2-like severity of Parkinson’s disease) receptor system [6]. • Differentiation between presynaptic par- kinsonism and other forms of parkinson- The most widely applied radiotracers for imaging D -like receptors with SPECT are ism (neuroleptic-induced parkinsonism) 2 123I-labelled IBZM and epidepride, both of Prior to the investigation, patients should which are commercially available. For PET, avoid taking any medications or drugs of [11C]raclopride, [18F] and [18F]des- abuse which could significantly influence methoxyfallypride (DMFP) are the most com- the visual and quantitative analysis of DAT monly used radiopharmaceuticals in Europe binding ligands. A withdrawal period of at [6]. These antagonist least 5 times the drug’s biological half-life is derivatives are not selective radiopharma- suggested. Smoking may interfere with DAT ceuticals for the D2 receptor, since they also availability [5]. bind the D3 receptor; however, the vast ma-

jority of D2-like receptors in the striatum are Brain neurotransmission SPECT/PET using D2 receptors. Since the available radiotracers dopamine D2 receptor ligands show considerable variation in their affinity The dopaminergic neurotransmitter system and selectivity for the D2 receptors and their plays a major role in movement disorders, pharmacokinetic properties, there are differ- particularly in parkinsonism. Using SPECT ences between them with respect to specific and PET, various functional aspects of dopa- binding ratios and the optimal time point minergic neurotransmission can be visual- for acquisition. As with DAT tracers, there is ised in vivo. physiological uptake of tracer in the striatum.

Currently nuclear medicine investigations According to the EANM Guidelines [6], com- predominantly assess two aspects of the mon indications for brain neurotransmission dopaminergic system: the binding of the SPECT/PET using dopamine D2 receptor li- presynaptic dopamine transporter and the gands are: status of the postsynaptic dopamine D2 re-

20 Chapter 2 Tracers for Brain Imaging

1. Differential diagnosis of parkinsonian neuropsychiatric disorders such as depres- syndromes (differentiation of Parkinson’s sion) [1]. The neuropharmacological data disease from other neurodegenerative obtained from receptor-specific PET/SPECT parkinsonian syndromes characterised by studies can additionally help to increase

loss of D2 receptors, i.e. multiple system knowledge of potential therapeutic targets atrophy and progressive supranuclear for novel pharmaceutical agents by deter- palsy) mining their dose-occupancy profile. This can be done using the radiolabelled drug under 2. Assessment of the extent of D receptor 2 investigation or by monitoring its effects on blockade during treatment with the binding of an established radioligand [1]. dopamine D2 antagonists (neuroleptics)

3. Huntington’s disease (D2 receptor imaging Tracers for the cholinergic system can be used

can confirm degeneration of postsynaptic for the study of neurodegenerative disorders EANM

D2 receptors) such as Alzheimer’s disease, Parkinson’s dis- ease, Lewy body dementia and progressive 4. Wilson’s disease (D receptor imaging 2 supranuclear palsy. Tracers for the central ben- findings are related to the severity of zodiazepine (BZD) receptors ([11C], neurological symptoms and may show [123I]) have been used in patients the degree of neuronal damage due to with epilepsy to identify the epileptogenic cytotoxic copper deposition in striatum) focus characterised by a focal decrease in

5. Pituitary adenoma (D2 receptor imaging BZD receptor density, for the study of neu- is helpful when assessing the dopamine ronal loss in Alzheimer’s disease and for the receptor status/availability of pituitary study of BZD receptor density in patients with adenoma, which may have implications or post-traumatic stress disor- for the medical treatment strategy) der (PTSD). Peripheral benzodiazepine bind- ing sites (PBBS) are present at low levels in the Tracers targeting other neurotransmitter normal brain. These sites are highly expressed systems and receptors in vivo by activated microglia, which are asso- Use of PET as a tool for neuroreceptor map- ciated with CNS inflammation in a wide range ping can be very important for elucidation of pathologies. In combination with MRI, to of basic mechanisms of disease and for in- aid with anatomical definition, PET imaging vestigation of correlations with clinical pa- using the PBBS-selective ligand [11C]PK11195 rameters. The antagonist radioligand [11C] provides a generic indicator of active disease WAY100635 has been used extensively to in the brain and, to date, has been used in study 5HT1A receptor (serotonin) dysfunction clinical studies of stroke, multiple sclerosis, in human disease states (neurodegenerative dementia, Parkinson’s disease, Huntington’s disorders such as Alzheimer’s disease and disease, epilepsy and schizophrenia [1].

21 Tracers targeting amyloid radiolabelled amino acids offer significant Amyloid plaques and neurofibrillary tangles improvements in the diagnostic evaluation are pathological markers found in the post of cerebral tumours. The contrast they dis- mortem of patients with Alzheimer’s play is far superior to that obtained with [18F] disease. It is thought that these plaques are FDG because of the low uptake of amino present as long as 10 years before any clinical acids in normal brain tissue, and they may symptoms of disease appear [1]. PET imag- be more tumour specific as their uptake ing of amyloid-β has emerged as a valuable is less influenced by inflammation [7]. The to support the in vivo diagnosis most frequently used radiolabelled amino of Alzheimer’s disease. Recently, several PET acid is (methyl-11C)-L- ([11C]MET). tracers have been developed that bind to In an effort to overcome the disadvantages amyloid plaques and hence can be used as of the short half-life and complex metabo- amyloid imaging agents (e.g. [11C]BIP, [11C] lism of [11C]MET, and despite changes in SB-13). [11C]BIP PET scans show a twofold in- amino acid structure, several fluoro- and crease in retention of signal in subjects with iodo-amino acids have been developed. Alzheimer’s disease compared with controls. These agents include 3-(123I)iodo-α-methyl-L- This increase in signal suggests widespread ([123I]IMT), used for SPECT, and O-(2- amyloid deposition in cortical areas and (18F)fluoroethyl)-L-tyrosine ([18F]FET), used striata in Alzheimer’s disease [1]. Further an- for PET. They are both transported by the alogues of the tracer labelled with 18F have same specific amino acid transport system as been developed and are currently being [11C]MET but are not incorporated into pro- introduced into clinical practice; examples teins. They display rapid accumulation into include [18F] (18F-AV-45), [18F]flute- brain tumours that is independent of BBB metamol and [18F] [1]. disruption (Fig. 5). Several 18F-labelled amino acids are available; [18F]FET, due to its ease of Brain tumour imaging using labelled synthesis, high in vivo stability and fast brain amino acid analogues and tumour uptake, has been selected as the Increased amino acid transport in brain tu- representative compound of this category. mour cells results from overexpression of the transporter systems and is related to al- terations in the tumour vasculature and tu- mour cell proliferation. In comparison with conventional anatomical imaging methods,

22 Chapter 2 Tracers for Brain Imaging EANM

A B

Figure 5 A, B: Brain PET/CT imaging with [18F]FET in a patient with suspected recurrence of glioblastoma. (A) PET image. An area of increased tracer uptake in the frontal lobe is seen on the right side. (B) CT image.

Despite differences in blood clearance, up- 2. Tumour delineation: Radiolabelled amino take kinetics and relation to protein syn- acid tracers are superior to CT and MRI for thesis, [11C]MET, [123I]IMT and [18F]FET show estimation of true tumour extension in similar results in the diagnostic evaluation low- as well as high-grade gliomas. Due to of cerebral tumours [7]. Doses in adults are lower uptake in normal brain tissue, they 100–400 MBq (typically 185 MBq) for [123I]IMT, are also superior to [18F]FDG for delinea- 200–250 MBq for [11C]MET and 200–250 MBq tion of low-grade tumours, in which [18F] for [18F]FET. FDG uptake is found to be decreased compared with normal cortex or basal According to the EANM Guidelines [7], com- ganglia. With [18F]FET, large brain vessels mon indications for brain tumour imaging may be visualised due to high blood pool using labelled amino acid analogues are: radioactivity. 3. Selection of the best site (labelled 1. Detection of viable tumour tissue amino acid imaging is recommended to (radiolabelled amino acid imaging is guide the stereotactic biopsy, which is superior to [18F]FDG-PET in confirming the gold standard in the classification and low-grade recurrence). grading of glioma).

23 4. Non-invasive tumour grading (grading As the number of PET/SPECT centres grows, of cerebral gliomas using labelled amino applications in clinical neurology will in- acids is controversial: [18F]FDG PET crease for early and/or presymptomatic di- appears better suited to differentiate agnosis of diseases. As more target-specific between tumour grades). radiotracers and ligands are developed, their use in clinical research and drug develop- 5. Therapy planning (in conjunction with ment will help determine optimal drug dos- anatomical imaging, radiolabelled amino ing regimens and elucidate the downstream acid imaging may be used to better de- effect of drug actions. fine the tumour volume to be resected or irradiated). 6. Tumour response (changes in uptake on labelled amino acid imaging may predict the response to locoregional chemo- and radiotherapy as early detection of residual tumour after surgery may be possible).

Acknowledgements. The author would like to thank Luka Lezaic for his valuable support.

24 2. (adapted from Sampson’s Textbook 4 ofRadiopharmacy, radiotracers/ligandsandtheirintended targets inclinicalneurology Common PET/SPECT Table 1 1. References References Chapter 2 4. 3. Amyloid 2Areceptors Serotonin (5HT) D Dopamine D Dopamine synthesis Dopamine receptors Serotonin (5HT) 1Areceptors Serotonin (5HT) Central benzodiazepine sites Peripheral benzodiazepine sites Dopamine reuptake (transporter) sites reuptake (transporter) Dopamine Cerebral bloodflow Target Cerebral glucosemetabolism Serotonin reuptake (transporter) sites Serotonin reuptake (transporter) Dierckx RAJO macy, 4 Theobald T, editor. Sampson’s Textbooks ofRadiophar Mol Imaging 2009;36:2103–10. Imaging Mol brain imaging using [18F]FDG, version 2. Eur J Nucl Med F, etal. Någren EANMprocedure K, guidelinesfor PET Varrone A, Asenbaum S, Vander Borght T, J, Booij Nobili 2009;36:2093–102. maceuticals, Imaging version Mol 2.Eur JNuclMed using99mTc-labelled SPECT perfusion radiophar T, etal. Någren EANMprocedure K, guidelinefor brain OL,NobiliF,Kapucu Varrone J, A,Booij Vander Borght New York: 2014. Springer; inneurology. Heidelberg andSPECT Berlin ders KL.PET th ed. London: Pharmaceutical Press; 2010. 2/3 1 , receptors Otte A, Otte receptors de Vries EFJde Vries , van WaardeA, Leen- - - 25 [ [ [ [ [ [ [ [ [ [ [ [ radiotracer/ligand PET/SPECT [ [ 11 11 11 11 11 11 11 18 18 15 11 123 11 11 5. 7. 6. F]fluorodeoxyglucose (FDG) O], C]diprenorphine [ C]PIB, C]flumazenil [ C]PK11195, C]DASB, [ 100907,[ C]MDL C]WAY 1000635 (11C)CFT, [ C]RTI-32, [ C]raclopride, C]SCH23390 F]fluorodopa (F-DOPA) I]epidepride Kapucu OL, et al. EANM procedureKapucu guidelines for brain J, J,Darcourt Booij Tatsch K, Varrone A, Vander Borght T, alogues. for braintumourimaging usinglabelledaminoacidan Ö,Kapucu Van etal. Laere EANMprocedure K, guidelines Vander Borght T, HalldinC, K, Asenbaum S,Bartenstein 2010;37:434–42. ceptor ligands, Imaging version Mol 2.Eur JNuclMed usingdopamineD2re neurotransmission SPECT/PET OL, et al. EANM procedureKapucu guidelines for brain Van Laere K, Varrone J, A, Booij Vander Borght T, Nobili F, ing 2010;37:443–50. ligands, Imag version Mol 2.Eurtransporter JNuclMed using123I-labelled dopamine neurotransmission SPECT 18 th F], [ ed. [1] 99m 123 Imaging Mol Eur JNuclMed I]β-CIT Tc-HMPAO, 123 11 I]PK11195 C]FLB457, [ C]FLB457, 18 F] 123 I]ioflupane 99m 123 Tc-ECD 123 I]IMPY I], . 2006;33:1374–80. - - - EANM Chapter 3 SPECT and SPECT/CT in Oncological Brain Imaging Elizabeth C. Hackett

Introduction the CT images. This results in a better image Single-photon emission computed tomog- showing both the physiological aspects of raphy, or SPECT as it is commonly known, the disease and anatomical information. It has been used for many years as an imaging also makes it much easier to pinpoint the tu- modality in the detection of oncological ab- mour location. normalities of the brain even though it is not the modality of choice for evaluation of brain Uses and advantages of SPECT and lesions. This chapter will discuss the evolu- SPECT/CT in brain oncology tion of SPECT and SPECT/CT imaging, the SPECT and SPECT/CT have been used widely uses and advantages of SPECT and SPECT/CT in the assessment of brain tumours and have in brain oncology, the radiopharmaceuticals been shown to be most useful in the imaging that have been used for this indication, imag- of recurrent gliomas and intracranial lympho- ing technique and data processing and, final- mas. SPECT imaging allows assessment of ly, the future of SPECT imaging in oncology. tumour malignancy based on the aspect, while SPECT/CT plays an im- The evolution of SPECT and SPECT/CT portant role in localisation of the tumour for SPECT imaging was introduced in the 1970s. the purpose of surgery. After surgery and/or It enables the creation of a 3D image from radiotherapy, SPECT and SPECT/CT are used multiple images taken by to help monitor the disease by evaluating cameras continuously rotating around an differences at the tumour site that can in- area of interest. The 3D images obtained dicate gliosis, radiation necrosis or tumour using SPECT offer improved sensitivity and recurrence. SPECT permits the functional im- resolution compared with 2D images and are aging of tumours of an indeterminate nature easier to compare with MRI and CT images. in order to evaluate their significance. It also SPECT also permits better tumour localisa- helps to differentiate intracranial tion as the 3D imaging component allows from toxoplasmosis. different views (sagittal, coronal and trans- axial) to be displayed, thereby offering im- Radiopharmaceuticals for brain imaging proved visualisation of the abnormality. with SPECT and SPECT/CT Several radiopharmaceuticals have been SPECT/CT cameras in which both the SPECT used for brain imaging in SPECT and gamma camera system and the CT camera SPECT/CT (Table 1). All of these radiopharma- system have been integrated into one gan- ceuticals are discussed in detail in Chapter 2, try have been available since the late 1990s. but we shall also cover them briefly in this The SPECT/CT systems allow for the fusion chapter. The most important quality that the of the nuclear medicine SPECT images with radiopharmaceutical must have is the ability

26 Chapter 3 SPECT and SPECT/CT in Oncological Brain Imaging

to cross the blood-brain barrier and localise Patient preparation: in normal brain tissue. This quality is needed • atients should avoid , alcohol and to permit evaluation of abnormalities within other drugs that can affect cerebral blood the brain tissue. Two of the most commonly flow for 24 h prior to the study. used radiopharmaceuticals are technetium- • Patients should not smoke cigarettes on 99m (99mTc) bicisate (ethyl cysteinate dimer, the day of the study. ECD) and 99mTc-exametazime (hexamethyl- propylene amine oxime, HMPAO). • A thorough detailed history needs to be taken, focussing on past head trauma and HMPAO is used in SPECT imaging in both un- previous radiological studies. stabilised and stabilised preparations. 99mTc- Pre-injection: HMPAO and 99mTc-ECD can appear to show • Ensure that the patient can comply with normal to decreased activity in abnormal EANM the scan procedures. brain tissue, but 99mTc-ECD has been known to show increased activity in tumours. Thal- • Explain the scan procedure to the patient lium-201 chloride (201Tl) is also used in SPECT or to the designated healthy care proxy or oncological imaging as it has high uptake caregiver. in brain tumours and can assist in biopsy • Have the patient relax in a dimly lit room 201 guidance. Figure 1 demonstrates Tl SPECT where there are no external stimuli, scans of different grades of glioma. including direct light, music, reading or speaking, for 20–30 min prior to the Technetium-99m sestamibi has also been injection. Whether the eyes are kept shown to have increased activity in tumours opened or closed depends on the and has been used in SPECT oncological im- individual department’s policy. aging. Figure 2 shows transaxial slices of a 99mTc-sestamibi SPECT scan. • Obtain intravenous access with either an angiocath or a butterfly needle at least 10 The injected activities and uptake times for min prior to injection. each of the above radiopharmaceuticals are • Instruct the patient not to move, speak or displayed in Table 1. read for at least 5 min after injection.

Imaging technique and data processing SPECT and SPECT/CT imaging should be performed bearing the following guidelines in mind:

27 • The patient should relax in the dimly • A camera with multiple detectors or lit room for 20–60 min or according to a dedicated brain imaging camera is the departmental procedure during the preferred to a single-head system for uptake phase of the radiopharmaceutical SPECT. A SPECT/CT camera is also preferred (see Table 1). as the scans can be fused for evaluation. Acquisition: • The energy peak should be set at 140 keV • After the uptake phase has been and the energy window to 20%. completed, the patient should void before • Low-energy high-resolution or ultra high- being brought into the camera room. resolution collimators should be used • Position the patient on the imaging and the highest resolution collimators table in a supine position so that he or are preferred. Fan beam, parallel-hole she is comfortable, with the arms down. collimators can be used if needed. Position the head so that it is in the • The matrix should be 128×128 or greater. middle of the field of view of the camera. Acquisition pixel size should be 1/3–1/2 The entire brain should be included in the the expected reconstructed resolution. field of view and a head holder or other It may be necessary to zoom in order to light head restraint should be used to achieve the required pixel size. help reduce patient motion. • The orbit should be 360° and as close as • When imaging, use the smallest radius possible to the patient. possible around the patient’s head to ensure diagnostic images are obtained. • A continuous acquisition is preferred It is ideal to be able to set up a contour over step and shoot with 20–40 s per radius to enable the camera head to be as projection. close as possible to the head at all times. • Segmenting the data will help reduce If the camera does not allow for a contour patient motion in the resulting images. radius to be set up, then it is best to use the smallest circulatory radius possible to Data processing: ensure maximum image resolution. • All studies should be filtered in three dimensions (x, y and z), using low-pass • Monitor the patient throughout the filters (Butterworth) and at the highest scan to ensure safety and to watch for pixel resolution. motion. • Be sure to reconstruct the entire brain from vertex to cerebellum using either the filtered back projection (FBP) or the iterative reconstruction method.

28 Chapter 3 SPECT and SPECT/CT in Oncological Brain Imaging

• Attenuation correction should be used Future of SPECT and SPECT/CT on all data unless a specific protocol or SPECT and SPECT/CT for imaging of brain case states otherwise. When a SPECT/CT tumours both entails challenges and of- is performed, the CT scan should be used fers further potential rewards. PET scanners for the attenuation correction. In the case have rapidly gained popularity in the field of a stand-alone SPECT camera, Chang of nuclear medicine as a means of obtain- attenuation correction should be used. ing functional images of brain tumours. In most cases, SPECT tumour studies provide • Data should be reformatted for display in a similar amount of diagnostic information coronal, sagittal and transverse slices. to that acquired with PET. This is useful as • When SPECT/CT is performed, not there are areas that do not have easy access only should the CT scan be used for to PET radiopharmaceuticals or PET cameras.

attenuation correction (as mentioned As stated earlier, the main drawback of SPECT EANM above) but the resulting SPECT images imaging of tumours is the lack of anatomical can also be fused with CT for detailed detail. However, the advent of SPECT/CT images. has allowed for the co-registration or fusion of scans, resulting in images that show both anatomical and functional detail. Such im- ages are very useful for tumour localisation and treatment planning.

29 References Chapter 3

1. ACR-SPR Practice Parameter for the Performance of Sin- 6. Mariani, G., et al. A Review on the Clinical Uses of SPECT/ gle Photon Emission Computed Tomography (SPECT) CT. Eur J Nucl Med Mol Imaging. 25 February 2010. Brain Perfusion and for Brain Death Examinations. 2014. 7. Juni, J. et. Al. Procedure Guideline for Brain Perfusion 2. Alazraki, N, Shumate, M., Kooby, D., A Clinican’s Guide SPECT Using Tc-99m Radiopharmaceuticals 3.0. Society to Nuclear Oncology: Practical Molecular Imaging and of Nuclear Medicine Inc. 2009. Therapies. Society of Nuclear Medicine, Inc. 2007. pp163-164 8. Schillaci, O., Filippi, L., Manni, C., Santoni, R. , Single Photon Emission Computed Tomography/Computed 3. Cherry, S., Sorenson, J., and Phelps, M., Physics in Nuclear Tomography in Brain Tumors. Semin Nucl Med. 2007 Medicine-3rd edition. W.B. Saunders Co. 2003. Jan;37(1); 34-47.

4. Chowdhury, FU., Scarsbrook, AF., The role of hybrid 9. State-of the-Art Nuclear Medicine: SPECT and SPECT/ SPECT-CT in Oncology: Current and Emerging Appli- CT. Association of Imaging Producers & Equipment cations. Clin Radiol. 2008 Mar; 63(3):241-51 Epub 2008 Suppliers. European Industrial Association for Nuclear Jan 14. Medicine and Molecular healthcare. 2009

5. Farrell, MB., et al. Quick Reference Protocol Manual for 10. Ziessman, H., O’Malley, J., and Thrall, J., Nuclear Medicine: NM Technologists. Society of Nuclear Medicine and Requisites-3rd edition. C.V. Mosby Co. 2005. Molecular Imaging. 2013. 122.

Table 1: Radiopharmaceuticals used for brain imaging in SPECT systems, with injected activities and uptake time

Injected Radiopharmaceutical Route Uptake time activity Recommended uptake time: 90 min 555–1110 MBq Intravenous 99mTc-HMPAO Minimum uptake time for interpretable (15–30 mCi) (through filter) images: 40 min Recommended uptake time: 45 min 555–1110 MBq 99mTc-ECD Intravenous Minimum uptake time for interpretable (15–30 mCi) images: 20 min 740 MBq Recommended uptake time: 99mTc-Sestamibi (20 mCi) Intravenous 20–30 min

Uptake time: 10–30 min 111–148 MBq 201Tl Intravenous Delayed images can be acquired (3–4 mCi) 2–4 h post injection

30 Chapter 3 SPECT and SPECT/CT in Oncological Brain Imaging

A EANM

B Tl SPECT scans demonstrating uptake in different grades of glioma: A grade 2 glioma; B grade 3 glioma; SPECTTl grade 2 glioma; B grades of glioma: A scans demonstrating uptake in different 201

C Figure 1A–C: Figure C grade 4 glioma. C grade © Neurographics and Dr. Thomas Booth Thomas and Dr. © Neurographics

31 Neurographics and Dr. Thomas Booth Thomas and Dr. Neurographics

Figure 2: 99mTc-Sestamibi SPECT scan of a brain glioma demonstrating uptake in the tumour.

32 18 18 so farare glucose metabolism by meansof main metabolicfeatures extensivelystudied ing to different clinicalindications.two The imag- also offers oftailoring thepossibility only tocontributes differential diagnosis, but not ofradiopharmaceuticals. PET/CT ability thanks to theincreasingly widespread avail- grownhas similarly especially inimportance, imaging ofbraintumours Combined PET/CT planning and post-therapeutic monitoring. sation, preoperative radiotherapy evaluation, toevant diagnosis,- classification,characteri ous metabolicfeatures ofgliomasthatare rel- fromto addresstance deriving - itsability vari agnostic work-up ofbraintumours, itsimpor of themostsuccessful techniques inthedi - isone positron emissiontomography (PET) intarget tissuesinvivo.cal activity Currently, andbiochemi- generate maps of functional imagingMolecular techniques are usedto Introduction Giorgio Testanera andGiovanna Pepe inOncologicalImaging BrainDiseases:PET/CT Chapter 4 ceuticals suchas - poration) usingamino acid radiopharma quantitative imaging. preparation; somenotes are alsoprovided on gether withacquisitionmethods andpatient clinical applicationsofthesetechniques, to on thetopic [1–5],we shallcover themajor guidelines withEANMandinternational ity thischapter,dopa). In ofcontinu- inthespirit dihydroxyphenylalanine ( 3'-deoxy-3'-fluorothymidine ( F-fluoroethyltyrosine ( (incor F-FDG and amino acid transport 11 C-methionine ( 18 F-FET), F-FET), 18 F-DOPA; fluoro 18 F-FLT) and 18 F-labelled 11 C-MET), C-MET), 18 F- - - - - 33 by the PET tomograph itself,by thePET usinga achieved by transmissionimaging acquired scanners, was attenuation correction PET brainimaging. oldergeneration In PET/CT for Attenuation ismandatory correction CT protocols data sets. aminations andenablesco-registration both the patientinsamepositionfor bothex oftheexaminationwith portions andCT PET lows sequentialacquisitionofcorresponding al- conjoined patient handling system [1]. It system withasingle, andaCT tion ofaPET technique thatemploysmodality acombina- isdefinedasanintegrated ormulti- PET/CT acquisition techniques PET/CT transmission sources with source. Alternatively, systems had somePET significantly reduces theradiationexposure. 10–30 rection, usingalow tubecurrent (typically only for ofattenuation cor thepurpose purposes, usinga regular tube current, or scancanbedonefor diagnostic tion. ACT oftheattenuation correc- ing theaccuracy of theradiopharmaceutical, withoutaffect theinjection after scan to be performed CT by the emission photons; thisallows the tion ofX-rays scanisnotaffected from theCT scan is that the detec- advantage of the CT scan for attenuation[6]. correction CT The systems have to usethe theability PET/CT mAs), i.e. a low-dose CT scan,which mAs), CT i.e. alow-dose 137 Cs. 68 Ge/ 68 Ga Ga - - - EANM The choice of type of CT scan depends on the PET protocols purpose of the imaging and the clinical indi- High-quality PET brain images can be ob- cations. If anatomical information is already tained only if strict procedures are adhered available, a low-dose CT scan for the purpose to, including respect to patient preparation, of attenuation correction can be considered, administration of adequate radiopharma- whereas if recent anatomical information is ceutical activity, patient positioning and use not available, a diagnostic CT scan (with con- of standardised acquisition protocols [8]. tiguous slices) may be preferred. A diagnos- tic CT (high mAs, contrast media injection) Patient preparation may vary for different requires specific skills in and spe- radiopharmaceuticals and different indica- cific radiologist guidelines must be followed. tions (Table 2). It is commonly accepted When CT is used for attenuation correction that for 18F-FDG PET the patient should fast and localisation only (i.e. is not intended to for at least 4 h to allow optimal cerebral be clinically diagnostic), the suggested work- uptake that is not influenced by increased flow is: serum glucose levels. Blood glucose levels should be checked prior to 18F-FDG admin- • CT topogram istration. When hyperglycaemia is present (>160 mg/dl), 18F-FDG uptake is reduced in • Low-dose CT the whole brain and stochastic noise is in- • PET acquisition creased. However, in brain tumours, hyper- glycaemia does not need to be corrected When a contrast-enhanced diagnostic CT scan and can even enhance detectability [1]. is also needed, it is advisable to perform it after PET acquisition, to avoid any possible Before the scanning procedure is started, the impact of IV contrast on PET images and patient should void the bladder to ensure standardised uptake value calculation. maximum comfort during the study. The pa- tient should be informed about the total ac- CT acquisition parameters (Table 1) are quisition time and advised to void again after strictly dependent on the equipment avail- the scanning session to minimise radiation able and national regulations, the latter be- exposure. The patient should be positioned ing particularly relevant to patient X-ray ex- comfortably in a quiet, dimly lit room for posure. The ALARA principle always applies several minutes before 18F-FDG administra- and must be followed to avoid unnecessary tion and during the uptake phase of 18F-FDG exposure. (at least 20 min) and should be instructed not to speak, read or be otherwise active. It

34 Chapter 4 Imaging in Oncological Brain Diseases: PET/CT

is desirable to have the cannula for intrave- In preparation for a radiolabelled amino nous administration in place 10 min before acid PET scan, patients are advised to take 18F-FDG administration. If a dynamic image a low-protein meal 4 h before the injection, is requested, the above procedures are to be although this requirement is controversial. followed with the patient positioned directly The patient may be requested to avoid tak- on the tomograph bed instead of in a differ- ing specific drugs during preceding clinical ent room. evaluations [10].

Post-processing routines allow correction for Imaging may be performed as a dynamic ac- minor obliquities of head orientation, so it is quisition or a static acquisition 20 min after more important to reduce the probability of injection of 330–400 MBq of 11C-methionine motion during acquisition than to achieve or 18F-FET. As with 18F-FDG, radiopharmaceu- perfect head alignment. The patient must be tical administration consists in intravenous EANM informed of the necessity to avoid voluntary injection of a bolus followed by flushing with movements of the head and must be asked physiological saline solution. After the ad- for her/his active cooperation. If cooperation ministration, an interval of 10 min must be is poor, sedation may be required. The pa- allowed before scanning. tient’s head should be only lightly restrained, avoiding use of a fixed container that can Table 3 shows examples of how PET/CT scan create claustrophobia. If movement artefacts is performed with two tomographs accred- can be expected, segmentation of data ac- ited by the EARL project [11–13]. As noted quisition into multiple sequential acquisitions above, we suggest that protocols should be may permit the exclusion of segments of pro- adapted according to the scanner employed, jection data affected by patient motion. following company suggestions and guide- lines, both national and international. The administered dose and scan param- eters are strictly dependent on national 18F-DOPA is an amino acid radiopharmaceu- regulations and the scanner used. EANM tical that accumulates in brain tumours via a guidelines for 18F-FDG [1] suggest, for transporter-mediated mechanism similar to adults, 300–600 MBq (typically 370 MBq) that responsible for uptake of methionine. in 2-D mode and 125–250 MBq (typically Many imaging protocols have been tested, 150 MBq) in 3-D mode. for instance with scanning starting at an

35 earlier time point after injection (e.g. 20 min) formance. The programme requires imaging or at a late time point, as is recommended sites to perform strict continuing quality con- for striatal imaging (e.g. 60–70 min) [14]. The trol, making them highly eligible as partici- European Association of Nuclear Medicine pants in multicentre studies. Participation in guidelines on paraganglioma imaging rec- the EARL programme is also useful as a qual- ommend a dose of 4 MBq/kg and imaging ity self-test for departments, to ensure that at 30–60 min after injection. In the evalua- routine patient examinations are of a high tion of brain tumours the best time interval quality [11–13]. for acquisition is between 10 and 30 min after injection because tumour uptake is For tumour imaging, semiquantitative esti- near maximum and because imaging at this mates of glucose metabolism such as the juncture is sufficiently early to avoid peak standardised uptake value (SUV) are typically uptake in the striatum. All these parameters used. SUVs may vary significantly if acquisi- refer to adults; we recommend the EANM tion procedures are not standardised, and paediatric dosage card as a useful tool for correct, reproducible standardisation re- determination of the activity to be adminis- quires efforts at every step of the schedule. tered in children. For 18F-FDG exam quantitation, a static image is sufficient, usually acquired at 30 or 60 min Quantitative PET/CT after injection (after 18F-FDG has reached a PET/CT scans offer the possibility of acquir- plateau concentration in the lesion). In ad- ing quantitative information on radiophar- dition, the exact total activity of 18F-FDG maceutical biodistribution inside a living administered and the patient’s weight and organism, but robust standardisation is re- height for measurement of body surface area quired to ensure reproducibility and to al- are required. A calibration factor is also need- low the use of imaging biomarkers in multi- ed. These semiquantitative estimates can be centre trials in which different scanners are corrected for blood glucose concentration. employed. Efforts aimed at harmonisation of acquisition, evaluation and quantification of When using methionine, the first quantifica- PET/CT studies have been concluded or are tion method routinely used is the tumour-to- ongoing worldwide. normal background ratio (T/N ratio), which compares uptake in the tumour to that in the In Europe, the research branch of EANM, contralateral frontal lobe or the correspond- EARL, set up the 18F-FDG PET/CT accredita- ing contralateral hemisphere. Generally, a tion programme, which is also endorsed by threshold greater than 1.5–1.9 is used for the the EORTC imaging group. It was launched diagnosis of brain tumour, but large prospec- in July 2010 with the objective of providing tive studies are needed to set a fixed T/N cut- a minimum standard of PET/CT scanner per- off ratio. SUVs should also be used (maximum

36 Chapter 4 Imaging in Oncological Brain Diseases: PET/CT

or mean) but their true diagnostic value is reconstruction and attenuation correction as still a matter of debate. MET accumulation are used in routine scans. Matching spatial in brain tumours reaches a plateau at 5–10 resolution is the most important parameter min after the radiopharmaceutical injection. for optimal database use. This allows assess- As for 18F-FDG scans, the SUV is influenced ment of normal variability of regional 18F-FDG by many factors, some being technical, such uptake and improves diagnostic accuracy. as acquisition starting time, and others more clinical, such as the plasma concentrations of Data interpretation should take into consid- N-acetylaspartate, which may affect the up- eration global changes, such as relative corti- take of MET in a competitive fashion. cal hypometabolism and regional decreases or increases in 18F-FDG uptake. Increased up- Quality criteria and artefacts take can be observed in active epileptogenic

Compliance with the above procedures may foci, tumours and inflammation. Known mor- EANM be expected to ensure an appropriate, sym- phological changes such as atrophy should metrical and readily interpretable represen- be considered in the interpretation. tation of the reconstructed dataset. Internal landmarks can be used for reorientation to The following list identifies some possible achieve a standardised image display. Reori- sources of misinterpretation that must be entation procedures based on the intercom- taken into consideration when deciding missural line are commonly used. The display whether a scan matches quality criteria: of additional coronal and sagittal images is mandatory. Three-dimensional display of • Artefacts (Figs. 1,2) the dataset can be helpful for more accu- --Patient movement rate topographic orientation in some clinical --Scanner related questions and to appreciate overall patterns --Induced by inappropriate processing of disease. The images should be critically ex- --Insufficient attenuation correction amined by technologists after the scan to en- --Soft tissue or skull uptake following sure that quality criteria are matched. Tech- surgery in the area of the skull or brain nologists also have to discuss with reporting • Unintended cerebral activation (i.e. visual physicians particular cases that may require or motor activation) additional scanning or reconstructions. Dur- ing interpretation for the presence of move- • Psychotropic drugs or corticosteroid use ment or attenuation artefacts, it is desirable • Sedation to have a normal database available, prefer- ably obtained on the same type of tomo- • Recent radiotherapy or chemotherapy graph, using the same acquisition position- ing and parameters and the same type of

37 Figure 1: Movement artefact: MIP (top left), PET (top right), CT (bottom left) and fused PET/CT (bottom right) images

38 Chapter 4 Imaging in Oncological Brain Diseases: PET/CT EANM

Figure 2: Metallic artefact: MIP (top left), PET (top right), CT (bottom left) and fused PET/CT (bottom right) images

Brain imaging with 18F-FDG background uptake of 18F-FDG in normal brain, The Warburg effect, i.e. the increase of an- which limits the possibility of discriminating aerobic even in aerobic conditions, neoplastic lesions, 18F-FDG PET/CT applications explains the glucose avidity of tumour cells are reported in the literature for the purposes and constitutes the basis for the 18F-FDG up- of tumour diagnosis, metabolic grading and take (Fig. 3). Despite the high physiological prognosis, volume estimation and follow-up.

39 Figure 3: 18F-FDG-positive brain lesion on axial CT (left) and fused PET/CT (right) images

Delbeke et al. described a sensitivity of 94% application that has been investigated is the and a specificity of 77% for18 F-FDG PET in use of 18F-FDG PET/CT to differentiate among the detection of high-grade gliomas [16]. common enhancing brain tumours such as Some authors suggest that delayed imag- lymphomas, high-grade gliomas and metas- ing at 3 up to 7 h post injection could help tases [20]. in the analysis of normal grey matter uptake and oncological uptake [17]. 18F-FDG PET/CT imaging is also used to provide an early assessment of the efficacy 18F-FDG uptake has been used to assess the of chemotherapy based on its value in pre- prognosis in patients with high-grade (grade dicting tumour metabolic response to te- 3 or 4) astrocytomas, the uptake being high- mozolomide, which allows oncologists to er in patients with a poorer prognosis and adopt a more tailored approach [21]. Some lower in those with a superior mean survival authors have studied the potential of 18F-FDG [18]. Similar results regarding the associa- to improve accuracy in stereotactic biopsy. tion of tumour 18F-FDG uptake with patient The literature available so far on the use of survival were also reported by Alavi et al.: a 18F-FDG PET/CT in primary tumours remains positive PET scan with hypermetabolic brain somewhat heterogeneous; nevertheless, lesions correlated with a poor prognosis as thanks also to its widespread diffusion and the rate of glycolysis increased in line with standardisation, this technique is still attrac- the degree of malignancy in primary cerebral tive and promising as a complementary pro- tumours [19]. Another interesting potential cedure that assists in completion of tumour

40 Chapter 4 Imaging in Oncological Brain Diseases: PET/CT

characterisation as the basis for improving or Other 18F-labelled radiopharmaceuticals changing patient management [22]. L-Dihydroxyphenylalanine (L-DOPA) is not an amino acid, but it is regarded as an amino Brain imaging beyond 18F-FDG acid analogue, being the product of the es- In order to overcome the limitations of the sential amino acid L-tyrosine. It was first in- use of 18F-FDG in brain imaging, which are troduced as a marker of the dopaminergic due to its physiological uptake in the grey system and is widely used to investigate the matter, other radiopharmaceuticals have in vivo transport to the basal ganglia for the been developed. In particular, radiola- diagnosis and follow-up of patients affected belled amino acids (11C-methionine, 18F-FET, by Parkinson’s disease. Within oncology, it 18F-DOPA) are currently the most commonly is known as a radiopharmaceutical for the used PET radiopharmaceuticals for imaging noradrenaline system, which may be used to of brain tumours. Another radiopharmaceu- investigate neuroendocrine tumours such as EANM tical introduced for imaging of cellular pro- phaeochromocytoma and neuroblastoma. liferation is choline, a precursor in the phos- pholipid synthesis required for the produc- 18F-DOPA PET/CT has also been proposed for tion of the cell membrane. Choline is usually the study of primary brain tumours because labelled with 11C or 18F. Somatostatin recep- of the high uptake in tumours (Fig. 4) as com- tor expression on the tumour surface, on the pared with the very low uptake in normal brain other hand, may be investigated by means of tissue (the basal ganglia represent the only ex- somatostatin receptor PET/CT thanks to the ception). A sensitivity of 85% and a specificity development of somatostatin analogues. of 89% have been reported for 18F-DOPA PET in the detection of primary brain tumours, and In this chapter, however, we prefer a more the method was identified as having further technical approach to a purely clinical one, prognostic potential as a predictor of tumour and we have therefore decided to group ra- grade and proliferation [24]. diopharmaceuticals according to the radio- nuclide used (Table 4) instead of the tumour features analysed. Therefore the following sections will focus on 18F- radiopharmaceuti- cals other than 18F-FDG, 11C-radiopharmaceu- ticals and 68Ga-radiopharmaceuticals.

With regard to the metabolic pathways and tumour features, the main families of avail- able radiopharmaceuticals are summarised in Table 5.

41 Magnetic resonance imaging (MRI) still rep- resents the mainstay in primary brain tumour imaging; however, in an interesting study in- volving comparative evaluation of 18F-DOPA PET and MRI in a population of 91 patients with histologically proven gliomas, concor- dance between MRI and PET was observed in 90% of cases [26].

The use of 18F-DOPA for detection of recur- rence of primary neoplastic disease in the brain is still a largely unexplored area.

18F-Fluoroethyltyrosine (FET) is an artificial amino acid taken up into upregulated tu- mor cells via enhanced expression of type L Figure 4: 18F-DOPA PET/CT showing increased amino acid carriers and increased carrier-me- uptake in a left temporo-mesial astrocytoma diated transport, but not incorporated into proteins (in contrast to natural amino acids While a standard acquisition protocol is usu- such as methionine). In a systematic review ally employed when imaging with 18F-DOPA, and meta-analysis by Dunet et al., 18F-FET in a study by Schiepers et al that focussed on showed an excellent performance in the ini- the kinetics of 18F-DOPA in brain tumours, dy- tial evaluation of newly diagnosed brain pri- namic PET images were acquired over a peri- mary tumours, with a sensitivity of 82% and od of 75 min starting soon after injection. The an average specificity of 76%.18 F-FET yielded results suggested that 18F-DOPA was trans- high-contrast images for both high- and low- ported but not trapped in tumours and that grade brain tumour, with low physiological the uptake curve could be related to tumour uptake in surrounding normal brain [27]. grade. High-grade tumours showed an early peak of activity followed by a sharp decline whereas low-grade tumours showed a more smoothed curve with a gradual decline [25].

42 Chapter 4 Imaging in Oncological Brain Diseases: PET/CT

18F-Fluorothymidine (FLT) is a marker of DNA synthesis in brain tumours; its uptake by brain tumours is possible in areas with a dis- rupted blood-brain barrier. Its sensitivity for low-grade gliomas has been shown to be lower than that of methionine PET.

11C-radiopharmaceuticals A Methionine is an amino acid that is usually labelled with 11C for brain tumour imaging. Owing to the many further evaluations be- yond the early feasibility studies, it is the most commonly used radiopharmaceutical for this EANM purpose in different countries. Analogously to 18F-FET, its increased uptake appears to be a consequence of the upregulation of L-type amino acid transporter 1 (LAT1), reflecting cell proliferation.

11C-methionine PET/CT is indicated both at B diagnosis, when it can contribute in assess- ing the degree of malignancy and the prog- Figure 5: Methionine PET/CT [sagittal (A) and nosis, in guiding biopsy and in evaluating coronal (B) views] showing multiple foci of tumour extent during radiation therapy (RT) recurrent glioblastoma planning, and after treatment, for evaluation of persistence of disease, response to therapy or relapse and for differentiation of recur- rence from necrosis (Figs. 5–7).

43 A B C

Figure 6 A–C: 18F-FDG PET/CT and 11C-methionine. An occipital brain lesion seen on CT images (A) shows faint 18F-FDG uptake (B) and intense 11C-methionine uptake

A B C

Figure 7 A–C: MRI (A), 18F-DOPA (B) and 11C-methionine (C) imaging of a low-grade astrocytoma

11C-methionine PET offers high-contrast im- Imaging can be performed as a dynamic or aging of the brain, reportedly superior to the static acquisition 10 min after injection of contrast achieved with 18F-FDG PET. It allows 370 MBq of 11C-methionine. the detection of both high- and low-grade gliomas, but it is especially useful with regard A multimodality approach including con- to the former, for which it has a sensitivity be- trast-enhanced MRI, 11C-methionine PET and tween 76% and 95% [28]. Other intracranial 18F-FDG-PET is suggested for tumour grading tumours such as benign and malignant me- and prognostication: a higher methionine ningiomas, pineal adenomas, and haeman- uptake is related to a poor outcome, albeit gioblastomas also show intense methionine better in low-grade and non-contrast-en- uptake, although false-positive results may hancing gliomas, whereas higher 18F-FDG be seen in benign conditions, such as de- uptake is a better predictor of poorer out- myelination, leukoencephalitis and abscess. come in high-grade and contrast-enhancing

44 Chapter 4 Imaging in Oncological Brain Diseases: PET/CT

gliomas. 11C-methionine PET has also been Towards the end of 1990s, 11C-choline was in- reported to detect progression with a sensi- troduced as a novel radiopharmaceutical to tivity of 90% and a specificity of 92%, whereas evaluate brain and prostate cancer [30] and MRI alone showed a lower specificity (50%) in 2001 brain tumour images were acquired for progression [29]. with 18F-choline in a patient with recurrent anaplastic astrocytoma [31]. The increased phosphorylcholine synthesis in tumours constitutes the substrate for the Ohtani et al. compared 11C-choline PET, MRI clinical application of the choline-based ra- and 18F-FDG PET in a series of patients with diopharmaceuticals in the imaging of brain suspected brain tumours. They observed cancer. The concentration of choline in nor- a correlation between 11C-choline and the mal cerebral cortex is low, whereas moder- histological tumour grade, with higher cho- ate uptake is seen in the choroid plexus and line uptake in high-grade gliomas, and sug- EANM cavernous sinus. Extracerebral tissues such gested that the combination of 11C-choline as extraocular eye, masticatory muscles and PET and MRI could improve the detection of bone marrow show a moderately higher ra- high-grade gliomas [32] (cf. Fig. 8). diopharmaceutical activity; parotid glands demonstrate intense uptake.

A B C

Figure 8: Oligodendroglioma. 11C-Choline PET (A), PET/CT (B) and post-processing fusion with brain MRI (C). The yellow arrow highlights the temporal lesion. Normal uptake is present in the choroid plexus In a comparative analysis of 11C-choline, appeared more useful in evaluating oligo- 11C-methionine and 18F-FDG PET, 11C-methio- dendroglial tumours. Both were superior to nine imaging was shown to be useful for the 18F-FDG PET in terms of tumour localisation evaluation of grade, type and proliferative in all types of glioma [33]. activity of astrocytic tumours, while choline 45 Glioblastomas and meningiomas are de- technique for radiation therapy planning, of- scribed to show moderately intense choline fering precise delineation of the biological uptake, whereas uptake in grade II and III target volume, and it is also able to discrimi- gliomas is generally faint. nate between radiation necrosis and tumour recurrence with higher accuracy than MRI or Choline PET appears a promising imaging 18F-FDG PET.

A B

Figure 9 A, B: 68Ga-DOTANOC imaging of a frontal meningioma

68Ga-radiopharmaceuticals ic tract; however, some intracranial tumours Somatostatin is a neuropeptide produced can show increased expression of SSR and by various neuronal, endocrine and immune therefore can be evaluated for this feature as cells. Imaging of somatostatin receptor (SSR) well. Contradictory data have been reported expression on tumour cells is widely applied on the positivity of 68Ga-DOTA-peptide PET and standardised. There are several peptides in series of glial tumours: Reubi et al. [34] with a different range of affinity for soma- concluded that SSRs are expressed predomi- tostatin receptor subtypes; at present the nantly by low-grade gliomas, with 82% of most frequently used radiopharmaceuticals low-grade gliomas (WHO grades 1 and 2) are 68Ga-DOTATOC, 68Ga-DOTANOC (Fig. 9) showing SSR expression compared with only and 68Ga-DOTATATE. These radiopharmaceu- 2% of high-grade gliomas (WHO grades 3 ticals are routinely employed for the imag- and 4). Others reported SSR expression ing of neuroendocrine tumours, in particular mainly in high-grade gliomas, relating this those arising from the gastroenteropancreat- observation to the loss of differentiation [35].

46 Chapter 4 Imaging in Oncological Brain Diseases: PET/CT

Meningiomas are the most common non- or CT but also to improve the contouring in glial brain tumour arising from the arach- RT planning and contribute in predicting the noid membrane and overexpress SSR, par- response to radiopeptide therapy. Other pos- ticularly the SSR subtype 2; this explains the sible applications of SSR PET with respect to utility of 68Ga-DOTA-peptide imaging for brain tumours are represented by pituitary meningioma [36]. adenoma, haemangioblastoma, medullo- blastoma and neuroendocrine metastases to In this scenario the role of SSR PET is to better the brain from other sites [37]. define lesions that are indeterminate at MRI

Acknowledgements. The authors would like to thank Sabrina Leonardi and all the technologists working in the Nuclear Medicine Department at Humanitas Research Hospital for their pre- cious help. EANM

47 References Chapter 4

References 1. Varrone A, Asenbaum S, Vander Borght T, Booij J, Nobili 11. Boellaard R, O’Doherty MJ, Weber WA, Mottaghy FM, F, Någren K, et al. EANM procedure guidelines for PET Lonsdale MN, Stroobants SG, et al. FDG PET and PET/CT: brain imaging using [18F]FDG, version 2. Eur J Nucl Med EANM procedure guidelines for tumour PET imaging: Mol Imaging. 2009;36:2103–10. version 1.0. Eur J Nucl Med Mol Imaging. 2010;37:181- 200. 2. Van den Broek WJM, Testanera G. Principles and practice of PET/CT. Part 2: A technologist’s guide. European As- 12. Boellaard R, Hristova I, Ettinger S, Sera T, Stroobants S, sociation of Nuclear Medicine. 2010. Chiti A, et al. EARL FDG-PET/CT accreditation program: Feasibility, overview and results of first 55 successfully 3. Vander Borght T, Asenbaum S, Bartenstein P, Halldin C, accredited sites. J Nucl Med. 2013;54 (Suppl 2):2052. Kapucu Ö, Van Laere K, et al. EANM procedure guidelines for brain tumour imaging using labelled amino acid an- 13. Boellaard R, Hristova I, Ettinger S, Stroobants S, Chiti A, alogues. Eur J Nucl Med Mol Imaging. 2006;33:1374–80. Bauer A, et al. Initial experience with the EANM accredi- tation procedure of FDG PET/CT devices. Eur J Cancer. 4. Gregoire V, Chiti A. PET in radiotherapy planning: Par- 2011;47 (Suppl 4):S8. ticularly exquisite test or pending and experimental tool? Radiother Oncol. 2010;96:275–6. 14. Antunovic L, Rodari M, Rossi P, Chiti A. Standardiza- tion and quantification in PET/CT imaging. PET Clin. 5. International Atomic Energy Agency. A guide to clinical 2014;9:259–66. PET in oncology: Improving clinical management of cancer patients. International Atomic Energy Agency. 15. Chondrogiannis S, Marzola MC, Rubello D. 18F-DO- 2008. PA PET/computed tomography imaging. PET Clin. 2014;9:307–21. 6. Sattler B, Lee JA, Lonsdale M, Coche E. PET/CT (and CT) instrumentation, image reconstruction and data 16. Delbeke D, Meyerowitz C, Lapidus RL, Maciunas RJ, transfer for radiotherapy planning. Radiother Oncol. Jennings MT, Moots PL, et al. Optimal cutoff levels of 2010;96:288–97. F-18 fluorodeoxyglucose uptake in the differentiation of low-grade from high-grade brain tumors with PET. 7. European Commission. European guidelines on qual- Radiology. 1995;195:47–52. ity criteria for computed tomography, EUR 16262EN. Luxembourg: Office for Official Publications of the Eu- 17. Spence AM, Muzi M, Mankoff DA, O’Sullivan SF, Link ropean Communities. 2000. JM, Lewellen TK, et al. 18F-FDG PET of gliomas at de- layed intervals: improved distinction between tumor 8. Schelbert HR, Hoh CK, Royal HD, Brown M, Dahlbom and normal gray matter. J Nucl Med. 2004;45:1653–9. MN, Dehdashti F, Wahl RL. Procedure guideline for tumor imaging using -18-FDG. J Nucl Med. 18. Patronas NJ, Di Chiro G, Kufta C, Bairamian D, Kornblith 1998;39:1302–5. PL, Simon R, Larson SM. Prediction of survival in glioma patients by means of positron emission tomography. J 9. Singhal T, Alavi A, Kim CK. Brain: Positron emission to- Neurosurg. 1985;62:816–22. mography tracers beyond [18F]fluorodeoxyglucose. PET Clin. 2014;9:267–76. 19. Alavi JB, Alavi A, Chawluk J, Kushner M, Powe J, Hickey W, Reivich M. Positron emission tomography in pa- 10. Moulin-Romsee G, D’Hondt E, de Groot T, Goffin, J Sciot tients with glioma. A predictor of prognosis. Cancer. R, Mortelmans L, et al. Non-invasive grading of brain 1988;62:1074–8. tumours using dynamic amino acid PET imaging: does it work for 11C-methionine? Eur J Nucl Med Mol Imag- 20. Kosaka N, Tsuchida T, Uematsu H, Kimura H, Okazawa ing. 2007;34:2082–7. H, Itoh H. 18F-FDG PET of common enhancing malig- nant brain tumors. Am J Roentgenol. 2008;190:W365–9.

48 21. References Chapter 4 24. 23. 22. 29. 28. 27. 26. 25. rakami M,etal. Cerebral withme rakami glioma:evaluation CL, Walter MA, etal. Correlation of6-18F-fluoro-L-dopa Fueger BJ, Czernin J, Cloughesy T, Silverman DH, Geist Imaging. 2012;56:173–90. yond FDGfor braintumorimaging. Mol QJNuclMed be radiopharmaceuticals Gulyás B, PET HalldinC.New Clin.2015;10:59–73. braintumors. PET primary fluorodeoxyglucosetomography PET/computed for EA,HessS, GrupeP,Segtnan PF. Høilund-Carlsen ¹⁸F- Biol Phys. 2006;66:331–8. not temozolomide plusradiotherapy. Oncol JRadiat Int grade gliomapredictsresponse to temozolomide but inpatientswithhigh- rate measured usingFDG-PET GM, N, Charnley West CM, CM, Barnett aminoacid. Q J Nucl Med Mol Imaging.aminoacid. 2012;56:151–62. Mol QJNuclMed F,Crippa withradiolabeled GL. PET AlessiA,Serafini thionine PET. Radiology. 1993;186:45–53. Ogawa T, ShishidoF, I, Kanno 2012;53:207–14. A systematic review andmetaanalysis. JNuclMed. for braintumor: thedifferential diagnosis ofprimary PET (18F-FET) mance of18F-fluoro-ethyl-tyrosine Dunet V, C,BuckA,StuppR,Prior Rossier JO. Perfor diol. 2009;71:242–8. - gliomas: initial experience. Eur J Ra primary/recurrent J,ernin etal. 18FFDOPA fusioninpatientswith PET/MRI Ledezma CJ, Chen W, Sai V, Freitas B, Cloughesy T, Cz- 2007;48:1651–61. SC. 18F-FDOPA inbraintumors. JNuclMed. kinetics C,Chen Schiepers W, Cloughesy T, DahlbomM,Huang 2010;51:1532–8. diagnosednewly andrecurrent gliomas. JNucl Med. withproliferation uptake andtumorgradePET in , et al. changeinglucosemetabolic Early Glaser M, et Inugami A, Inugami Brock C, Fujita H, Bydder Bydder Mu - - - - 49 30. 36. 35. 34. 33. 32. 31. 37. rol. 2004;63:13–9. astrocytomas. J Neuropatholand high-grade Exp Neu- and sst3somatostatin receptor proteins inlow-grade etzmann K S,Pauli C,Schulz Mawrin SU, system. Cancer Res. 1987;47:5758–64. nervous matostatin receptors intumorsofthehumancentral ofso andbiochemicalcharacterization Distribution JC,Lang Reubi W, JW, Maurer R,Koper SWJ. Lamberts J Neuroradiol. 2008;29:1176–82. tomography. positron-emission 11C-choline Am AJNR ing 11C-methionine, [18F]fluorodeoxyglucose, and A, Kato T, N, ShinodaJ, Nakayama hanced MR imaging. Eur J Nucl Med. 2001;28:1664–70. andgadoliniumen withFDGPET choline: comparison Ohtani T, H,IshiuchiS, Kurihara 2001;42:1805–14. tracers. JNuclMed.choline analogsasoncologic PET man HS,etal. of[18]F-labeled Synthesis andevaluation DeGrado TR, BaldwinSW, Wang S, 1997;38:842–7. JNuclMed.of braintumorwith[methyl-11C]choline. Hara T, N,ShinouraKondo T. Kosaka imaging PET peptides? AJR AmJRoentgenol. 2013;201:1340-7. peptides? AJR tumors: apotential area ofapplicationfor 68Ga-DOTA ofintracranial Somatostatin receptor-based PET/CT P,Sharma R. A, Kumar A, Bal C, Malhotra Mukherjee 2001;42:1053–6. results inpatientswithmeningiomas. JNuclMed. tors using[68Ga]DOTAd-Phe1-Tyr3-octreotide: first DW Henze J, P, M, Schuhmacher Hipp oue T Yano H,etal. assessmentofgliomasus Metabolic , Doll J Doll , Sasaki T Sasaki , et al. PET imaging ofsomatostatin al. recep PET , et , etal. Differential expression ofsst1,sst2A, . Braintumourimaging withcarbon-11 Treuheit T Saito N, Orr MD Orr Miwa K Miwa Kowalski J Kowalski , Oriuchi N, Oriuchi Liao RP , Diete S, , Okumura Okumura , Becker Becker , Fried Di In ------EANM Tables Table 1: Recommended preparation, contrast medium injection and CT acquisition parameters for brain imaging (adapted from [7])

Patient preparation Diet (3 h)

Patient position Supine/head first

Contrast medium injection Contrast medium 300–350 mg l/ml

First injection 60 ml, 1 ml/s

Delay 3 min

Second injection 50 ml, 1.5 ml/s

Acquisition Starting at the end of second injection

CT acquisition parameters Volume of investigation Vertex to C1

Collimation N x 0.75 mm

Slice thickness 3–5 mm

Slice increments Continuous

Pitch 1

Rotation time 0.75–1 s

mAs 250

X-ray tube voltage 120

Resolution matrix 512 x 512

FOV 240 mm

Reconstruction filter Soft tissue

50 6-slice CT; scanner B is from 2012, with time-of-flight technology and 64-slice CT (adapted from [14]) technology and64-slice scannerBisfrom 2012,withtime-of-flight 6-slice CT; technology and EARL-accredited scanners:scannerAisfrom 2006,withnotime-of-flight PET/CT Table 3:Practical examplesofdifferent parameters used twofor brainimaging different with from [9]) Table 2:Patient preparation otherthan whenusingradiopharmaceuticals References Chapter 4 18 18 18 11 11 Topogram protocol CT parameters modulation Dose Slice Collimation Rotation time Rotation Pitch attenuation correction for Reconstruction imaging for Reconstruction C-Methionine C-Flumanezil F-Fluoroethyltyrosine F-DOPA F-Fluorothymidine Scanner A Scanner 256 mm 110 kV 50 mA 130 kV 80 mA 3 mm 6 x2mm 1 s 1.0 3 mm, FOV 300mm 3 mm,FOV smooth+ H10s very + 3 mm, FOV 250mm + 3mm,FOV H31s mediumsmooth Low-protein diet4hbefore scan(nostandard/common practice) (no standard/common practice) No seizure 1day before during scan Low-protein diet4hbefore scan(nostandard/common practice) No dopaminergic ordopaminomimeticdrugs12hbefore scan No fastingrequired 51 Topogram CT protocol CT parameters modulation Dose Slice Collimation Rotation time Rotation Pitch attenuation correction for Reconstruction imaging for Reconstruction Scanner B Scanner 250 mm 120 kV 10 mA 140 kV value 280 mA,maximum 3.75 mm 64 x3.75mm 1 s 0.984:1 3.75 mm, FOV 500mm 3.75 mm,FOV AC wideview 0.625 mm, FOV 500 mm 0.625 mm,FOV Standard wideview 18 F-FDG (adapted

EANM Scanner A Scanner B

PET protocol PET protocol

Scan duration bed 10 Scan duration bed 10 min

Matrix 256 Matrix 256

Zoom 2 DFov 30 cm

Reconstruction Iterations 2, subsets 8 Reconstruction VUE Point FX, iterations 3, cut-off 4 mm, subsets 18, sharp IRON

Filter Gaussian Filter No filter axis Z

FWHM 4 mm FWHM /

Abbreviations: AC, attenuation correction; DFOV, display field of view; FOV, field of view; FWHM, full-width at half-maximum; IRON, General Electric proprietary name; VUE Point FX, General Electric proprietary name

Table 4: PET in brain tumour imaging (modified from [23])

Radionuclide Half- Mean ß+ Maximum ß+ Example of radiopharmaceutical life energy energy (min) (MeV) (MeV)

18F-Fluorodeoxyglucose (FDG) 18F-Dihydroxyphenylalanine (F-DOPA) 18F 109 0.25 0.63 18F-Fluoroethyltyrosine (FET) 18F-Fluorothymidine (FLT) (18F-Choline)

11C 20.4 0.39 0.96 11C-Choline 11C-Methionine

68Ga 68.1 0.84 1.90 68Ga-DOTA-peptide

52 Table familiesinbraintumourimaging radiopharmaceutical 5:PET References Chapter 4 Somatostatin receptor expression Phospholipids: cellmembranesynthesis Nucleosides: DNAreplication/cell proliferation Amino acidanalogues Amino acids:protein synthesis pathways/tumourMetabolic features 53 68 ( 11 18 18 18 11 Radiopharmaceutical 18 Ga-DOTA-peptides C-Choline F-Fluorothymidine (FLT) F-Dihydroxyphenylalanine (F-DOPA) F-Fluoroethyltyrosine (FET) C-Methionine F-Choline)

EANM Chapter 5 Imaging in neurological and vascular brain diseases (SPECT and SPECT/CT) Filipa Lucena, Eva Sousa and Tânia F. Vaz

Introduction dementia, movement disorders, traumatic Since the first in vivo studies of cerebral func- brain injury, cerebrovascular diseases (acute tion with radionuclides by Ingvar and Lassen stroke, chronic ischaemia, preoperative eval- [1], nuclear medicine (NM) brain applications uation and assessment of brain death), brain have evolved dramatically, with marked im- tumours and brain infections [7–10]. provements in both methods and tracers. Consequently it is now possible to assess not The available radiopharmaceuticals include only cerebral blood flow and energy metab- brain perfusion tracers and neurotransmis- olism but also neurotransmission [2]. Planar sion binding agents. There are currently no functional imaging was soon substituted by SPECT tracers for the specific study of brain single-photon emission computed tomogra- metabolism [11] (cf. Chapter 2). phy (SPECT) and positron emission tomog- raphy (PET); it now has limited application in Brain perfusion brain imaging, being reserved for the assess- Brain perfusion scintigraphy allows the as- ment of brain death [3, 4] sessment of regional cerebral blood flow (rCBF), using radiopharmaceuticals that dis- In recent years, owing to their wide availability tribute proportionally to the regional tissue and superior spatial resolution, structural tech- blood flow after intravenous injection and niques such as computed tomography (CT) remain trapped for a sufficient length of time and magnetic resonance imaging (MRI) have to allow acquisition of SPECT images. These replaced NM functional imaging techniques specific perfusion compounds are small, neu- in the assessment of patients with neurologi- tral, lipophilic and able to cross the blood- cal disorders. However, SPECT and PET can brain barrier by passive diffusion [12]. The make valuable contributions in an expanding most widely used radiopharmaceuticals for number of clinical situations, complementing brain perfusion SPECT are 99mTc-hexameth- structural imaging [5, 6]. In fact, multimodal ylpropylene amine oxime (99mTc-HMPAO) approaches now have an established clinical and 99mTc-ethyl cysteinate dimer (99mTc-ECD). role in neuroimaging applications, whereby Non-specific perfusion radiopharmaceuti- the structural images permit anatomical lo- cals, such as 99mTc-diethylenetriamine penta- calisation of the radiopharmaceutical distribu- acetic acid (99mTc-DTPA), can also be used for tion and also allow for some corrections (e.g. the evaluation of brain death, providing in- attenuation correction) [6]. formation on intracranial vascular flow.

Commonly accepted clinical indications for Patient considerations SPECT brain imaging include presurgical General information regarding patient prep- lateralisation and localisation of epilepto- aration can be found in Table 1, including genic foci and the evaluation of suspected specific considerations for each step of the

54 Chapter 5 Imaging in neurological and vascular brain diseases (SPECT and SPECT/CT)

procedure. From a general point of view, Correct head positioning should be achieved patient preparation includes the withdrawal using a head holder provided by the gamma of any factor known to disturb rCBF, as it will camera manufacturer [13]: the centre of the affect the normal pattern of radiopharma- head of the patient should be aligned with ceutical distribution in brain. The length of the centre of rotation, preventing displace- time for which restrictions are imposed var- ment of the head to the periphery of the field ies depending on the duration of the distur- of view; the canthomeatal line (i.e. the line bance produced in the rCBF. The restrictions extending from the ear to the eye) should be may include dietary constraints, withholding oriented as vertically as possible; and there of medication and control of environmental should be no rotation or lateral tilt of the variables, such as exposure to visual, auditory head. Commercially available brain SPECT and somatosensory stimuli. Special attention processing applications allow reorientation should be paid to potential disturbances in of the data after acquisition; however, these EANM the minutes prior to injection of the radio- data manipulations may introduce errors, in- pharmaceutical and during the uptake phase. cluding image blurring, highlighting the fact that head positioning is an important step in The preparation procedure requires correct the procedure [13, 14] evaluation of the patient’s ability to cooper- ate and good communication skills on the Immobilisation devices should be used to part of the NM technologists so that all the provide slight restraint and, when comple- information and recommendations can be mented by a comfortable position for the effectively conveyed [7, 13]. In uncoopera- patient, reduce the probability of movement tive patients, sedative medication (e.g. short- during SPECT acquisition [13]. acting benzodiazepine) can be considered as long as administration is after the radiophar- Different acquisition protocols (cf. Table 1) maceutical uptake phase, thereby avoiding can be used according to the clinical indica- sedation-induced blood flow changes [7, 8]. tions. Currently, planar imaging is used only for brain death studies, in early and delayed Technical considerations regarding image phases, either alone or in combination with acquisition and processing SPECT, with brain-specific or non-specific When positioning the patient for image ac- agents. On the other hand, SPECT imaging quisition, NM technologists should observe is acquired in most clinical situations, after various guidelines in order to avoid factors intravenous injection of 370–1110 MBq (in that may interfere with the obtained data adults) of 99mTc-brain-specific radiopharma- and the clinical interpretation of results ceuticals and with a variable time delay be- (cf. Table 1). tween the injection and the acquisition in ac- cordance with the radiopharmaceutical and

55 the clinical indication [7, 8, 15, 16]. Regardless is simpler and preferable. The challenge por- of these variations, images should be com- tion is performed first and if this yields normal pleted within 4 h after injection owing to the results, omission of the baseline study may be [7]. considered. For further information regard- ing this intervention, the relevant guidelines Concerning SPECT acquisition parameters, should be consulted [7, 8]. the radius of rotation should be as short as possible, preferably less than 15 cm [7, 13, 17]. · Epileptic focus localisation: It is well estab- In this context, head holders deliver an addi- lished that during an epileptic seizure an in- tional advantage in eliminating the distance crease in cortical blood flow occurs in the area imposed by the shoulders and gamma cam- of the focal discharge [11]. Patients with focal era bed and allowing closer positioning of the epilepsy refractory to therapy may benefit detectors around the patient’s head. A typical from surgical ablation of the seizure focus [4] SPECT acquisition is best acquired over a 360° and brain perfusion scintigraphy contributes rotation, along 120 projections with sufficient in localisation of the epileptogenic focus dur- time per image to obtain a total number of ing the ictal state [11]. For this purpose a 2-day counts exceeding 5 million (without scatter protocol that documents perfusion in the ictal correction) [7, 13, 17]. Multidetector or dedi- and interictal states can be performed. The cated brain SPECT cameras will contribute main challenge when using this procedure to a decrease in total scan time and yield concerns the radiopharmaceutical injection. high-quality images. Also, detector sensitivity Individual doses must be readily available at should be taken into account so that the time the patient’s bedside, allowing for rapid injec- per projection can be controlled in order to tion by a trained NM technologist at the onset obtain suitable statistical data [7, 13, 17]. of the seizure. In this way, the distribution of the radiopharmaceutical will reflect the rCBF Occasionally, brain perfusion scintigraphy at the time of ictus and delayed SPECT images can be performed under the effect of certain will show the location of increased blood flow. interventions with the goal of performing Interictal SPECT studies add useful informa- comparisons with a baseline procedure. In tion to the ictal study [7, 8]. these cases, acquisition and processing are identical to those for the general procedure: General information on technical consider- ations regarding imaging acquisition and · Vasodilatory challenge with acetazolamide processing is presented in Table 1. (DiamoxTM) or an equivalent is indicated in the evaluation of several brain vascular disorders as it leads to an increase in rCBF in normal ce- rebral vessels via dilatation. The 2-day protocol

56 Chapter 5 Imaging in neurological and vascular brain diseases (SPECT and SPECT/CT)

Table 1: General information and procedures for brain perfusion SPECT and SPECT/CT [3, 7, 8, 18, 19]

PATIENT PREPARATION

Before arrival • Avoid, if possible, caffeine, alcohol, smoking, and any drugs known to affect rCBF.

• Evaluate the patient for ability to cooperate (e.g. lie still for approximately 30–60 min). • Place an intravenous cannula 10–15 min prior to radiopharmaceutical injection. • Position the patient comfortably in a quiet room, with no interaction with other patients. • Instruct the patient to keep the eyes and the ears open and also to not speak, read At NM or move from at least 5 min before until 5 min after injection. department • If sedative medication is considered, it should be administered at least 5 min after

radiopharmaceutical injection, and a few minutes before data acquisition. EANM • If vasodilation is considered, contraindications must be evaluated and the dosage of acetazolamide, prepared. The radiopharmaceutical is injected about 15–20 min after injection of acetazolamide because it is at this time that the vasodilatory effect is most pronounced. • Patients should void prior to acquisition for maximum comfort during the study.

Continuous supervision of patients during the procedures is necessary, especially for those with epilepsy and dementing disorders. In the case of brain death scintigraphy no preparation is needed.

57 RADIOPHARMACEUTICAL

Adults: • 259–925 MBq for 99mTc-HMPAO (typical activity: 740 MBq) [7, 15] • 370–1110 MBq for 99mTc-ECD (typical activity: 1110 MBq) [8, 16] Administered Children (according to the EANM paediatric dosage card table Version 1.2.2014): activity • 99mTc-HMPAO: “baseline activity” = 51.8 MBq • 99mTc-ECD: “baseline activity” = 32 MBq • Administered activity = “baseline activity” × multiple • Minimum recommended: 110 MBq

DATA ACQUISITION

Time delay, • 99mTc-HMPAO: best image quality at 90 min; after 30 min the image will be injection - interpretable 99m imaging • Tc-ECD: best image quality at 30–60 min • Acquisition completed within 4 h after injection • Supine position with head on head holder and lightly restrained by immobilisation device Patient • Maximum possible vertical orientation of canthomeatal line positioning • No rotation or lateral tilt of the patient’s head • Head of the patient centred in the gamma camera centre of rotation Gamma • Multiple detector (triple or dual head) or dedicated brain SPECT camera • LEHR or LEUHR parallel-hole collimators. Fan-beam collimators might be preferred Photopeak • 15–20% energy window centred around 140 keV

Dynamica • 1–3 s per frame for at least 60 s. Matrix 64×64 pixels. Anterior projection • 500–1000 kcts. Matrix ≥128×128 pixels. Anterior, lateral and (if possible) posterior Statica projections • Radius <15 cm, Matrix ≥128×128 pixels, <3° angulations in a 360° rotation; hardware zoom may be necessary; mode acquisition step and shoot (frequently) or SPECT continuous (reduce scan times); 120 projections, 20–30 s per projection, >5 million total counts (without scatter correction) • Selected to give an acquisition pixel size 1/3 to 1/2 of the expected resolution Zoom (e.g. 1.5 zoom)

58 Chapter 5 Imaging in neurological and vascular brain diseases (SPECT and SPECT/CT)

DATA ACQUISITION

CT scan (SPECT/ • CT for attenuation correction: 120–140 kV, 2.5 mA, rotational speed of 1 s, slice CT systems) thicknesses of 3–5 mm, increment of 2.5 mm [20–23]

DATA PROCESSING

Review of • Review of projection data in cine mode and sinograms is helpful for an initial projection determination of scan quality, patient motion and artefacts. data • Motion correction algorithms, if available, may be used before reconstruction for minor movements, but rescanning is necessary if there is substantial head motion.

• Continuous colour scales. (For example, a common colour scale used is EANM represented here: . The brain regions with high Colour scale/ perfusion are orange to white, and regions with lower perfusion are blue to black.) table • After use of a standard colour scale, only adjustments for background subtraction or contrast are recommended.

• Reconstruct the whole brain, including cerebellum. • Filtered back-projection (FBP) or iterative reconstruction (e.g. OSEM) (may improve SPECT detection accuracy). Reconstruc- • Highest pixel resolution (one-pixel slice thickness). tion • Filter and reconstruction parameters (cut-off, order, iterations and subsets) will depend on the injected activity, camera sensitivity and resolution, type of acquisition, clinical application and diagnostic reporting sensitivity.

• Chang’s method (homogeneous correction matrix, μ=0.12–0.14 cm−1) using the Attenuation shape contouring of the head. correction • Non-uniform attenuation correction method (e.g. CT scan) is recommended.

• Images are reformatted into slices in 3 orthogonal planes (axial, coronal Display and sagittal). Correct reorientation is important for visual interpretation and semiquantification.

59 DATA PROCESSING

• Regions of interest (ROIs) can be used to determine tracer uptake ratios in different cerebral regions, allowing estimation of the relative rCBF distribution within the brain. Corresponding structures in contralateral hemisphere or another reference Semiquantifi- region (e.g. cerebellum, hemisphere, total brain) can be compared and asymmetry cation indices may be used to enhance differences in rCBF. • Spatial normalisation, comparison to normal databases and voxel-based analysis can also be performed. • Image subtraction can be useful in the assessment of patients with epilepsy. a Acquired only in cases of brain death scintigraphy with 99mTc-DTPA

Normal and pathological brain patterns specific scalp and facial tissue background ac- Although the brain presents a high vascular tivity) that seem to be related to the trapping demand (about 15–20% of the cardiac out- mechanism [11, 13, 17] (Fig. 1). put), its limited ability to store energy means that a tightly regulated blood flow is required in order to provide the necessary amount of glucose and oxygen [5, 17, 24, 25]. Synaptic activity at the neuronal cell bodies represents the primary utilisation of blood, explaining the increased grey matter to white matter Courtesy of Nuclear Medicine Department – CUF Descobertas, Lisbon, Portugal blood flow ratio of approximately 4:1 (80 ml/ Figure 1: Normal brain perfusion SPECT with min/100 g and 20 ml/min/100 g, respectively) 99mTc-HMPAO (transverse plane), showing [5, 11, 25]. This factor impacts on the distribu- cortical and subcortical grey matter uptake. tion of the radiopharmaceutical in brain struc- tures and the effect is visible on the acquired In several pathologies, this normal pattern images. Cortical and subcortical grey matter can be altered and the characterisation of structures are clearly distinguishable from distinct pathological patterns may help in white matter, presenting symmetrical uptake the differential diagnosis. For example, cer- in both hemispheres [13]. Even though both tain subtypes of dementia, such as Alzheim- 99mTc-labelled radiopharmaceuticals (99mTc- er’s disease (AD), frontotemporal dementia HMPAO and 99mTc-ECD) express the same spe- (FTD) and dementia with Lewy bodies (DLB) cific brain pattern, slight differences may be present different patterns of regional hypo- observed (e.g. with regard to regional uptake perfusion on brain perfusion scintigraphy. in the thalamus and the cerebellum and non- Classically, hypoperfused regions involve the

60 Chapter 5 Imaging in neurological and vascular brain diseases (SPECT and SPECT/CT)

parieto-occipital cortex in AD, the bilateral in predicting the malignant course of brain frontal cortex in FTD and both frontal regions swelling with large hemispheric infarctions and the posterior association cortex in DLB and in evaluating cerebral perfusion in non- [26]. acute cerebrovascular disease [18, 28].

In patients with intractable partial epilepsy Flow imaging of rCBF with suitable radio- who are candidates for epilepsy surgery, pharmaceuticals has been used for early brain perfusion scintigraphy has an impor- detection of stroke. Depending on the stage tant role in localising the ictal onset zone and of stroke, i.e. early or late (cf. Fig. 2), the scinti- seizure propagation pathways, both of which graphic fi ndings can be used to select those present as a focus or regions of hyperperfu- patients who may benefi t from thrombolytic sion on ictal phase imaging. Regions of func- therapy. Although CT and MRI are more of- tional defi cit associated with the ictal onset ten employed, SPECT is sometimes advanta- EANM zone may also be identifi ed on interictal geous owing to its ability to detect ischaemic phase imaging. In this context, image sub- events earlier and with greater sensitivity and traction may be a valuable tool in the assess- to predict accurately the infarct size (includ- ment and interpretation of the results [27]. ing the penumbra area, i.e. the area that ben- efi ts from early reperfusion interventions and Brain perfusion scintigraphy is also helpful is at the highest risk for expansion of irrevers- in vascular brain diseases owing to its value ible injury) [6, 18, 29].

rCBF is reduced Early stages “Misery (i.e. few hours) Tissue oxygen metabolism perfusion” is maintained Stroke Ischaemia progresses “Luxury Later stages perfusion” Damage to blood vessels

Figure 2: Early and later stages of stroke and fl ow/metabolism characterisation

Brain perfusion scintigraphy has also been using a cerebrovascular dilator such as ac- shown to be eff ective in the study of transient etazolamide. Since abnormal vessels do not ischaemic attacks. In these cases, imaging display a dilatory response to acetazolamide, is performed under pharmacological stress SPECT images show a signifi cant reduction in

61 radiotracer uptake in the cerebral regions fed with parkinsonism. Nowadays, by using sin- by the affected blood vessels (analogous to gle-photon emission tomography (SPET. observations when performing myocardial perfusion imaging with dipyridamole) [6]. Studies with radiopharmaceuticals that bind to transporters or receptors of neurotrans- In cases of chronic ischaemia, brain perfu- mitters are usually very accurate and offer a sion scintigraphy can assess the functional good approach for quantification of the do- reserve capacity before and after the ad- paminergic neurons lost. These techniques ministration of acetazolamide in order to of functional molecular imaging are very demonstrate impaired vascular reserve and useful for evaluation of disease progression maximum vasodilation in a territory referable and response to therapeutic interventions to the presenting symptoms [7, 19]. and for assessment and monitoring of the neural mechanisms underlying parkinson- Semiquantification based on regions of in- isms. Additionally, other major clinical ap- terest (ROIs) provide an estimate of relative plications are the differential diagnosis differences in rCBF based on comparison of between PD and other neurodegenerative radiopharmaceutical uptake ratios between parkinsonian syndromes, between parkin- affected and reference regions (e.g. right/left sonian syndromes and essential tremors asymmetries) [7, 19]. (ET) and between DLB and other dementias [9, 10, 30, 32]performing, interpreting and Brain neurotransmission SPECT reporting the results of clinical dopamine There are several neural pathways that can D2 receptor SPECT or PET studies, and to be evaluated by brain SPECT imaging. In this achieve a high quality standard of dopamine chapter we focus on the assessment of dopa- D2 receptor imaging, which will increase minergic system neurotransmission. the impact of this technique in neurological practice.The present document is an update Parkinson’s disease (PD) is a neurodegen- of the first guidelines for SPECT using D2 re- erative disorder in which there is progres- ceptor ligands labelled with (123. In Europe, sive loss of dopamine neurons in the nigros- the most widely used radiopharmaceuticals triatal system, leading to widespread motor for SPECT in this context are -123 symptoms (e.g. bradykinesia, rigidity, tremor N-(3-fluoropropyl)-2β-carbomethoxy-3β- and impaired balance) and cognitive impair- (4-iodophenyl)nortropane ([123I]FP-CIT) and ments [30, 31]including Parkinson’s disease, iodine-123 2β-carboxymethoxy-3β-(4-iodo­ multiple system atrophy and progressive su- phenyl) ([123I]β-CIT) for imaging of pranuclear palsy. The results of post-mortem the DAT (presynaptic) and [123I]iodobenza- studies point to dysfunction of the dopami- mide (123I-IBZM) and [123I]epidepride for imag- nergic neurotransmitter system in patients ing mainly of the D2 receptor (postsynaptic).

62 Chapter 5 Imaging in neurological and vascular brain diseases (SPECT and SPECT/CT)

The study of transporter and receptor bind- nergic neurotransmitter system in patients ing allows assessment of (a) the presynaptic with parkinsonism. Nowadays, by using sin- nigrostriatal neurons through evaluation of gle-photon emission tomography (SPET. the dopamine transporter (DAT) and (b) the postsynaptic dopaminergic function through Patient considerations evaluation of the dopamine receptors D2 and General information regarding patient prep-

D1 [9, 10, 30]including Parkinson’s disease, aration can be found in Table 2, including multiple system atrophy and progressive su- specific considerations for each step of the pranuclear palsy. The results of post-mortem procedure. studies point to dysfunction of the dopami-

Table 2: General information and procedures for brain SPECT with DAT ligands ([123I]β-CIT and

123 123 123 EANM [ I]FP-CIT) and dopamine D2 receptor ligands ([ I]IBZM and [ I]epidepride) [9, 10]

PATIENT PREPARATION

[123I]β-CIT and [123I]FP-CIT [123I]IBZM and [123I]epidepride

Before arrival Avoid smoking, and any drug that Withdrawal of drugs known to affect 2D causes binding potential deficit, such receptor binding, such as: dopamine as: cocaine, , central agonists, neuroleptics, other drugs (e.g. nervous system stimulants, , , cinnarizine, , , adrenergic agonists, , ). anticholinergic drugs, , anaesthetics.

• Evaluate the patient for ability to cooperate (e.g. lie still for approximately 30–60 min). At NM • For exams with 123I-ligands it is necessary to block the thyroid gland with an department adequate regimen (e.g. at least 200 mg of sodium perchlorate given at least 5 min prior to injection) to prevent accumulation of free radioactive iodine in the thyroid. • Patients should void prior to acquisition for maximum comfort during the study.

RADIOPHARMACEUTICAL

Administered • Adults: 150–250 MBq (typical activity, 185 MBq) [9, 10] activity • Children: No data available and no clinical indications established

63 DATA ACQUISITION

Time delay, • [123I]FP-CIT: 3–6 h after injection injection – • [123I]β-CIT: 18–24 h after injection imaging • [123I]IBZM: 1.5–3 h (best image quality at 2 h) • [123I]Epidepride: 2–3 h

Positioning • Similar to brain perfusion SPECT

The preparation procedure requires correct Technical considerations regarding imaging evaluation of the patient’s ability to cooper- acquisition, processing and quality control ate and good communication skills on the When positioning the patient for imaging part of the NM technologist so that all the in- acquisition, NM technologists should take formation and recommendations can be ef- into account various factors that may inter- fectively conveyed. Sedation can be consid- fere with the data obtained and the clinical ered in uncooperative patients and should interpretation of the results. Correct head po- be given at the earliest 1 h prior to the SPECT sitioning is essential, as previously described acquisition [9, 10]. for brain perfusion SPECT, and immobilisa- tion devices are also recommended [9, 10]. For exams with 123I-ligands it is necessary to block the thyroid gland with an adequate Different acquisition protocols (cf. Table 2) regimen (cf. Table 2). are used for image acquisition with DAT 123 123 ([ I]β-CIT and [ I]FP-CIT) and D2 receptor Patient preparation includes the withdrawal ligands ([123I]IBZM and [123I]epidepride). All of any drugs that affect the binding to DAT of these radiopharmaceuticals are admin- or D2 receptors: a withdrawal period at least istered by intravenous injection of 150–250 five times the drug’s biological half-life is rec- MBq (typically 185 MBq). For DAT tracers, the ommended. Antiparkinsonian drugs do not interval between the injection and image markedly affect DAT binding and therefore do acquisition is 8–24 h for [123I]β-CIT and 2–6 123 not need to be withdrawn prior to DAT SPECT, h for [ I]FP-CIT. Regarding the D2 receptor but it is possible that L-DOPA drugs down- tracers, the images should be acquired 1.5–3 regulate DAT expression and caution is ad- h (preferably 2 h) after administration of [123I] vised in intra-individual follow-up studies. The IBZM and 2–3 h after administration of [123I] binding of the radioligand to the D2 receptors epidepride [9, 10]. can be compromised by many antiparkinso- nian drugs (in particular dopamine agonists), neuroleptics and other drugs [9, 10].

64 Chapter 5 Imaging in neurological and vascular brain diseases (SPECT and SPECT/CT)

Regardless of the variations, it is recom- General information on imaging acquisition mended that a fixed interval is used between and processing is presented in Table 3. radiopharmaceutical administration and im- age acquisition to ensure that data are com- parable between subjects and intra-individu- al follow-up studies [9, 10].

Table 3. General data acquisition and processing information for brain SPECT and SPECT/CT 123 123 with dopamine transporter ligands ([ I]β-CIT and [ I]FP-CIT) and the dopamine D2 receptor ligand ([123I]IBZM and [123I]epidepride) [9, 10, 33, 34]

[123I]β-CIT and [123I]FP-CIT [123I]IBZM and [123I]epidepride EANM Gamma • Multiple detector (triple or dual head) or dedicated brain SPECT camera • LEHR or LEUHR parallel-hole collimators. Fan-beam collimators may be preferred • Smallest possible rotational radius (typical 11–15 cm) Rotation with appropriate patient safeguard • Angle increments of 3° (360° rotation) with circular orbit Matrix • ≥128×128 pixels Photopeak • 15–20% energy window centred around 159 keV • Selected to give an acquisition pixel size 1/3 to 1/2 of the expected resolution Zoom (e.g. 1.5 zoom) Acquisition • Step and shoot (frequently) or continuous mode (reduces scan times) • >1.5 million (FP-CIT); • >3 million Total counts >1 million (β-CIT) • Triple-head camera: 30 min (e.g. 120 • Dual-head camera: 40–50 min projections, 40 each head; (e.g. 120 projections, 60 each head; DATA ACQUISITION DATA 45 s/projection) 40–50 s/projection) Total scan • Dual-head camera: 30–45 min • Triple-head camera: for the same scan time (e.g. 120 projections, 60 each head; time, higher count statistics 30–40 s/projection) • The number of s/projection depends on the sensitivity of the system CT scan • CT for attenuation correction: 120–140 kV, 2.5 mA, rotational speed of 1 s, (SPECT/CT slice thicknesses of 3–5 mm, increment of 2.5 mm systems) • Anatomical location of radiopharmaceutical accumulation: [20–23] 20–40 mA (other parameters are the same)

65 [123I]β-CIT and [123I]FP-CIT [123I]IBZM and [123I]epidepride

• Review of projection data in cine mode and sinograms is helpful for an initial Review of determination of scan quality, patient motion and artefacts. projection • Motion correction algorithms, if available, may be used before reconstruction for data minor movements, but rescanning is necessary if there is substantial head motion. • Continuous colour scales, preferably the “cold” colour scales. Colour scale/ • After use of a standard colour scale, only adjustments for background table subtraction or contrast are recommended. • Filtered back-projection (FBP), or preferably, iterative reconstruction (e.g. OSEM). • For FBP, a low-pass Butterworth filter is recommended, with a cut-off of 0.4–0.5 Re­ and a power factor of 8–10. Filtering includes either 2D prefiltering of the construction projection data or 3D postfiltering of the reconstructed data. • For OSEM, the reconstruction parameters can be: 2, 4 or 6 iterations, 10 subsets. • The reconstructed pixel size should be 3.5–4.5 mm with one-pixel slice thickness. • Chang’s method (homogeneous correction matrix, μ=0.1 cm-1) using shape Attenuation contouring (if available) of the head. correction • CT scan (measured correction matrix). DATA PROCESSING DATA • Images are reformatted into slices in 3 orthogonal planes (axial, coronal and sagittal). Display Correct reorientation is important for visual interpretation and semiquantification. • Select transverse slices (highest striatal binding slices, at least 3 consecutive slices, or the entire striatal volume) for ROI definition. • ROIs should be drawn in the striatum and the striatal subregions (caudate, putamen), to assess specific dopamine transporter (CIT)/receptor ([123I]IBZM) Semiquanti- binding. The reference regions (background) without (or with low) density can fication be the occipital cortex or cerebellum. • Semiquantification techniques: classic manual ROIs (most widely used), manual volumes of interest (VOIs), more advanced automated systems using VOIs and voxel-based mathematical systems.

The influence of SPECT acquisition param- might require repetition of the acquisition. eters, such as radius of rotation, rotation, Minor patient movement can be easily cor- number of projections and use of dedicated rected using motion correction algorithms cameras, has already been described and provided by commercial workstations. In those considerations also apply to DAT and these exams a certain degree of movement

D2 receptor ligand scintigraphy. After the is to be tolerated given the patients’ clini- data acquisition, planar projection images cal conditions; nevertheless, the correction and the sinogram/linogram should be im- provided by algorithms may not be accurate mediately inspected by the NM technologist and therefore should be used wisely. in order to identify any patient motion that

66 Chapter 5 Imaging in neurological and vascular brain diseases (SPECT and SPECT/CT)

In general, after processing and image re- tential (low density of DAT and D2 ligands). formatting into three orthogonal planes, vi- The segmentation method is similar for DAT sual assessment of the images is performed and D2 receptor ligands. It is helpful if the and segmentation procedures are used to ROI size (which should be at least twice the obtain semiquantitative information. Sev- full-width at half-maximum) and shape are eral segmentation methods may be applied standardised (e.g. by use of an atlas or tem- (e.g. ROI) manually, semi-automatically or full plates). When feasible, ROI definition may be automatically. When segmentation is per- based on individual morphology as revealed formed manually, the use of non-continuous by image fusion with MRI, which is particu- colour tables may overestimate findings larly important when low specific binding is due to abrupt colour changes. Semiquanti- expected (e.g. in the case of a severe loss or fication is intended to allow assessment of blockage of the DAT) [9, 10, 35]. the ratio between specific binding (in the EANM striatum and striatal subregions) and non- It is useful to compare the patient-specific specific binding regions (e.g. occipital cor- binding values (cf. Eq.1) [33] with data from tex, cerebellum). The reference ROI is drawn normal subjects (preferably age-matched) in regions with absent (or low) binding po- obtained with the same technique.

(mean counts of striatal ROI – mean counts of non-specific ROI Specific binding values= (mean counts of non-specific ROI) (Eq.1)

These values can be obtained from a central Normal and pathological brain patterns database or can be established for each de- Functional imaging provides information on partment. It is only possible to use control regional abnormalities in dopaminergic sys- values from a central database if phantom tem function (pre- and postsynaptically). The studies are performed and are demonstrat- highest density of dopamine neurons oc- ed to allow comparable calculations for the curs in the nigrostriatal dopamine pathway, different imaging set-ups used. Sometimes which is related to the increased radiophar- intra-individual comparisons are useful in maceutical uptake by these structures in DAT patient follow-up (i.e. baseline vs follow-up and D2 receptor imaging. for therapy control or assessment of disease progression), and standardised evaluations In the transverse plane of the normal brain im- using approaches based on stereotactic nor- age, the striatum is clearly visible and symmet- malisation are especially useful for more reli- rical (right and left striatum). It is composed of able verification of subtle changes [9, 10, 35]. the caudate and the putamen, the head of the

67 caudate having the shape of a small sphere achieve a high quality standard of dopamine and the putamen, that of a crescent (Fig. 3). D2 receptor imaging, which will increase The image appearance is similar with both the impact of this technique in neurological types of ligand in healthy controls [30, 35]. practice.The present document is an update of the first guidelines for SPECT using D2 re- ceptor ligands labelled with (123, allowing the differential diagnosis between parkin- sonian syndromes and ET and also DLB and other dementias.

In the case of D2 receptor imaging, there is a difference in binding between PD and other parkinsonian syndromes, which allows dif- ferential diagnosis that is not possible on the basis of DAT imaging [30]including Parkin- (Courtesy of Nuclear Medicine Department – CUF Descobertas, Lisbon, Portugal) son’s disease, multiple system atrophy and Figure 3: [123I]FP-CIT scintigraphy (transverse progressive supranuclear palsy. The results of plane) reconstructed by FBP, showing the post-mortem studies point to dysfunction of symmetrical striatum in a healthy control. the dopaminergic neurotransmitter system in patients with parkinsonism. Nowadays, When imaging with DAT ligands, binding by using single-photon emission tomogra- decreases with the evolution of PD, parkin- phy (SPET. In the early stages of PD there is sonian syndromes (e.g. multiple system atro- a short-lived increase in D2 receptor binding phy and progressive supranuclear palsy) and (postsynaptic) in response to depletion of DLB. Initially, reduced uptake is observed in synaptic dopamine levels. Disease progres- the putamen, while in later stages caudate sion and increasing degeneration of nigros- uptake is also affected. In patients with PD, triatal dopamine neurons leads to a slight the uptake of DAT ligands is related to the decrease in binding in late stages; by com- clinical stage and disease severity and has an parison, in other parkinsonian syndromes a important role in early diagnosis and assess- decrease is observed throughout the course ment [30, 35]. [30]including Parkinson’s disease, multiple system atrophy and progressive supranucle- In dementias other than DLB and ET, the ar palsy. The results of post-mortem studies SPECT image with DAT ligands is normal point to dysfunction of the dopaminergic [9, 10, 30, 32]performing, interpreting and neurotransmitter system in patients with reporting the results of clinical dopamine parkinsonism. Nowadays, by using single- D2 receptor SPECT or PET studies, and to photon emission tomography (SPET).

68 Chapter 5 Imaging in neurological and vascular brain diseases (SPECT and SPECT/CT)

As the presynaptic dopaminergic neurons with CT (from SPECT/CT) rather than Chang’s are affected in PD and in atypical parkinson- method given that application of the lat- ism, it is useful to perform joint evaluation of ter approach often causes overestimation pre- and postsynaptic function, by assessing, of frontal perfusion and underestimation of with specific radiopharmaceuticals, DAT and posterior perfusion, i.e. non-uniform attenua-

D2 receptors [36]. tion correction is required to accurately esti- mate rCBF [20, 39]. On the other hand, when

Challenges and opportunities in SPECT performing DAT and D2 receptor imaging no and SPECT/CT imaging for neurological significant differences in visual interpretation and vascular brain diseases or semiquantification are observed depend- In recent years neuroimaging has evolved ing on whether CT or Chang’s method is used rapidly, with major contributions from func- for AC; Chang’s method is to be preferred in tional and structural imaging techniques. The this context in order to avoid the additional EANM greatest challenges at present include the radiation exposure from CT [21]. optimisation of instrumentation and com- putation and also the development of new, Crucial to the growth of NM techniques is the more accurate radiopharmaceuticals. continued development of molecular probes that can bind to the target biological recep- Improvements in image quality and quanti- tor with high selectivity. In the field of brain fication methods will enhance spatial reso- imaging, the development of new tracers lution and sensitivity, optimise the integra- with more specific and better characteristics tion of different modalities and support the is contributing greatly to improvements in development of new algorithms for recon- SPECT imaging, and numerous relevant re- struction and processing of data [37]. In this search initiatives are being conducted [40]. context the use of databases of normalised It is certain that new challenges and oppor- values representative of normal or pathologi- tunities will arise from closer collaboration cal populations also has an important role, between those working on fundamental/ providing anatomical information or localisa- applied research and on the clinical appli- tion and allowing more accurate quantifica- cations of brain SPECT; the outcome will be tion procedures [38]. Attenuation correction enhanced non-invasive in vivo visualisation (AC) for brain SPECT imaging is also a sub- of biological processes at the molecular and ject that will need further research. For brain cellular levels, to the benefit of patients. perfusion images, AC should be performed

69 References Chapter 5

References 1. Ingvar D, Lassen N. Quantitative determination 13. Morano GN, Seibyl JP. Technical overview of brain SPECT of regional cerebral blood-flow in man. Lancet. imaging: improving acquisition and processing of data. 1961;278:806–7. J Nucl Med Technol. 2003;31:191–5; quiz 202–3.

2. Baert AL, Sartor K, Schiepers C, eds. Diagnostic nuclear 14. Covington MF, McMillan NA, Avery RJ, Kuo PH. The semi- nedicine. Berlin, Heidelberg New York: Springer; 2006. colon sign: dopamine transporter imaging artifact from head tilt. J Nucl Med Technol. 2013;41:105–7. 3. Donohoe KJ, Agrawal G, Frey KA, Gerbaudo VH, Mariani G, Nagel JS, et al. SNM practice guideline for brain death 15. GE Healthcare. CERETECTM kit for the preparation of scintigraphy 2.0. J Nucl Med Technol. 2012;40:198–203. technetium Tc99m exametazime injection.

4. Mettler FA, Guiberteau MJ. Essentials of nuclear medi- 16. Lantheus . Neurolite manufacturer cine imaging, 6th edn. Philadelphia: Elsevier Health Sci- leaflet 7-18-2008. ences; 2012. 17. Catafau AM. Brain SPECT in clinical practice. Part I: Perfu- 5. Sharp PF, Gemmell HG, Murray AD, eds. Practical nuclear sion. J Nucl Med. 2001;42:259–71. medicine. London: Springer; 2005. 18. Masdeu JC, Irimia P, Asenbaum S, Bogousslavsky J, Brain- 6. Ahmadzadehfar H, Biersack H-J, eds. Clinical applica- in M, Chabriat H, et al. EFNS guideline on neuroimaging tions of SPECT-CT. Berlin, Heidelberg New York: Springer; in acute stroke. Report of an EFNS task force. Eur J Neurol 2014. Off J Eur Fed Neurol Soc. 2006;13:1271–83.

7. Kapucu OL, Nobili F, Varrone A, Booij J, Vander Borght 19. Latchaw RE, Yonas H, Hunter GJ, Yuh WTC, Ueda T, T, Någren K, et al. EANM procedure guideline for brain Sorensen AG, et al. Guidelines and recommendations perfusion SPECT using 99mTc-labelled radiophar- for perfusion imaging in cerebral ischemia. A scientific maceuticals, version 2. Eur J Nucl Med Mol Imaging. statement for healthcare professionals by the Writing 2009;36:2093–102. Group on Perfusion Imaging, From the Council on Car- diovascular Radiology of the American Heart Associa- 8. Juni JE, Waxman AD, Devous MD, Tikofsky RS, Ichise tion. Stroke. 2003;34:1084–1104. M, Heertum RLV, et al. Procedure guideline for brain perfusion SPECT using 99mTc radiopharmaceuticals 20. Bieńkiewicz M, Górska-Chrzastek M, Siennicki J, Gajos 3.0. J Nucl Med Technol. 2009;37:191–5. A, Bogucki A, Mochecka-Thoelke A, et al. Impact of CT based attenuation correction on quantitative assess- 9. Van Laere K, Varrone A, Booij J, Vander Borght T, Nobili F, ment of DaTSCAN ((123)I-Ioflupane) imaging in diag- Kapucu OL, et al. EANM procedure guidelines for brain nosis of extrapyramidal diseases. Nucl Med Rev Cent neurotransmission SPECT/PET using dopamine D2 re- East Eur. 2008;11:53–8. ceptor ligands, version 2. Eur J Nucl Med Mol Imaging. 2010;37:434–42. 21. Lange C, Seese A, Schwarzenböck S, Steinhoff K, Umland-Seidler B, Krause BJ, et al. CT-based attenu- 10. Darcourt J, Booij J, Tatsch K, Varrone A, Vander Borght T, ation correction in I-123-ioflupane SPECT. PloS One. Kapucu OL, et al. EANM procedure guidelines for brain 2014;9:e108328. neurotransmission SPECT using (123)I-labelled dopa- mine transporter ligands, version 2. Eur J Nucl Med Mol 22. Patton JA, Turkington TG. SPECT/CT physical princi- Imaging. 2010;37:443–50. ples and attenuation correction. J Nucl Med Technol. 2008;36:1–10. 11. Elgazzar AH, ed. The pathophysiologic basis of nuclear medicine, 3rd edn. Berlin, Heidelberg New York: Spring- 23. Buck AK, Nekolla S, Ziegler S, Beer A, Krause BJ, Her- er; 2015. rmann K, et al. SPECT/CT. J Nucl Med Off Publ Soc Nucl Med. 2008;49:1305–19. 12. Saha GB. Fundamentals of nuclear pharmacy, 6th edn. New York: Springer; 2010.

70 34. 34. 33. 32. 31. 28. 27. 26. 24. References Chapter 5 25. 30. 29.

Goffin K, Dedeurwaerdere S, Dedeurwaerdere K, Goffin K, VanLaere Van Paesschen Biersack H-J, Freeman LM. Clinical nuclear medicine. Park E. A new era of clinical dopamine transporter imag- Park eraofclinicaldopaminetransporter E.Anew Siennicki J, Bieńkiewicz M, Płachcińska A,GranjaC,Leroy M,Płachcińska J, Bieńkiewicz Siennicki Rodriguez-Oroz MC, Jahanshahi M, Krack P, M, Krack MC,Jahanshahi Rodriguez-Oroz I, Litvan in ofmedialtemporal patterns impairment Distinct Basely M,Ceccaldi M,Boyer O, L,Mundler GuedjE. neuronal B, Göbel Oltmanns KM,ChungM.Linking Djang DSW,Djang J, N,Booij MJR,Bohnen Janssen Henderson degenerative dementia: a brain SPECT perfusion study perfusion degenerative dementia:abrainSPECT Model. 2013;10:50. Med to theglucosemetabolism. brain activity Theor Biol content/aip/proceeding/aipcp/10.1063/1.3295652 http://scitation.aip.org/ing. AIPConf Proc 2010:1204. processing technique for [123]I−DaTSCAN Neuroimag acquisitionand C. OptimalmethodologyofSPECT∕CT Med.Soc Nucl 2012;53:154–63. Publ Off 1.0. JNuclMed imaging with123I-ioflupaneSPECT mine transporter TA, etal. guideline for SNMpractice dopa- K, Herholz 8. ing using 123I-FP-CIT. J Nucl Med Technol. 2012;40:222– mechanisms. LancetNeurol. 2009;8:1128–39. of Parkinson’s disease: features and pathophysiological R,Bezard clinicalmanifestationsMacias E,etal. Initial J NuclMed. 1999;26:171–82. Eur emission tomography inpatientswithparkinsonism. emissiontomographyusing single-photon andpositron 2003;22:327–35. prognostic Nucl. EspMed assessmentofstroke].Rev inthe dera E,Romero SPECT J, etal. [Brain perfusion Heidelberg New Berlin York: 2008. Springer; NuclMed. 2008;38:227–39. Semin W. Neuronuclear assessmentofpatientswithepilepsy. Imaging. 2013;40:932–42. Mol Eur JNuclMed in Alzheimer’s diseaseandfrontotemporal dementia. Nucl Med. 2014;55:1831–41. Heiss W-D. imaging inischemicstroke. Radionuclide J aging of the dopaminergic neurotransmission system J,Booij Tissingh G, - Winogrodzka A, van Royen EA. Im Serena A, Nogueiras JM,Outomuro A,Corre J, Ortega - - 71 40. 38. 36. 35. 39. 37.

Tatsch ofthedopaminergic Imaging system indif K. Badiavas K, Molyvda E,Iakovou I, Molyvda Badiavas K, Tsolaki M,Psarrakos Varrone A,DicksonJC, Tossici-Bolt L,Sera T, Asenbaum S, PET and SPECT imaging of disease. Rev. Chem Soc and SPECT PET Pimlott tracersfor A.Molecular the SL,Sutherland 2011;40:149–62. Nucl Med. 2014;39:e343–5. using(99m)Tc-ECD. SPECT/CT Clin in brainperfusion ofnonuniform CT-basedimpact attenuation correction Farid Petras K, S,Poullias X,Caillat-Vigneron N.Clinical 2013;40:213–27. ent methodsofanalysis.Imaging. Mol Eur JNuclMed effects, genderdifferences evaluationofdiffer and (ENC-DAT):controls age-related for [123I]FP-CIT SPECT J,Booij et al. European multicentre database of healthy 207. tions andchallenges. Commun. 2008;29:193– NuclMed strengths, versus SPECT: limita- A,Zaidi H.PET Rahmim Imaging. 2008;35Suppl1:S51–7. ferential diagnosis Mol ofdementia.Eur JNuclMed Imaging. 2011;38:764–73. Mol disorders: farbeyond visualassessment.Eur JNuclMed imaging inmovement evaluation N.SPECT Karatzas K, - - EANM Chapter 6 Imaging in Neurological and Vascular Brain Diseases (PET/CT) Ian Law, Valentina Garibotto and Marco Pagani

PET/CT cameras and brain imaging PET/CT clinical indications in neurological Since the introduction of the first systems and vascular diseases in 2001, the field of clinical PET/CT has ex- Dementias perienced impressive growth in terms of Dementia is a syndrome characterised by de- system design and functionality, supportive terioration in multiple cognitive functions, as- software solutions and the availability and sociated with functional impairment and with range of clinically approved PET tracers. De- a chronic course, due to a brain dysfunction. velopments were initially directed towards Neurodegenerative dementias are increas- the creation of solutions for implementation ingly prevalent, given their strong association primarily in systemic oncology, rather than with aging of the population, and are one of in brain studies. Over the intervening years, the most relevant causes of disability and de- however, clinical brain PET imaging has ben- pendency among elderly people [1]. Alzheim- efitted tremendously through the use of er’s disease (AD) is the most common form, ac- available equipment and organisation. The counting for up to 70% of cases; among other number of clinical brain PET scans at a clinical types, the most frequent are vascular dementia, site might not initially be sufficient to justify dementia with Lewy bodies (DLB) and fronto- the acquisition of a PET/CT scanner, but as a temporal lobar degeneration (FTLD). The neu- supplementary element in a production line rodegenerative process of dementia presum- dominated by oncology, brain PET studies ably lasts over decades, with a long preclinical can play a very significant role. phase without symptoms and a “prodromal” phase, identified by the term “mild cognitive There are no technical or practical limitations, impairment” (MCI), during which symptoms as such, to the potential uses of PET/CT scan- are mild, with a cognitive performance below ners for clinical brain PET scanning. Unen- average but without functional impairment hanced CT of the head is performed initially [2]. The differential diagnosis among the vari- and may be either of a clinical quality for di- ous forms is a critical and complex process, agnostic reading or a low-dose CT scan for particularly in these early phases. An early di- the purpose of attenuation correction. The agnosis is important for disease management CT scanning has a duration of <1 min, so the and proper treatment and also represents a re- number of CT slices, 4–128, is of no practical lief for patients and caregivers [3]. The different consequence in clinical routine. The number forms of dementia affect specific brain areas of CT slices does not significantly impact the and neurotransmission systems and have dif- clinical CT quality. All PET/CT systems have ferent neuropathological features, identified an axial field of view of 15–21 cm and will, by PET imaging, such as regional neuronal thus, be able to acquire a full 3D brain vol- dysfunction or abnormal protein depositions; ume in one bed position (see “Fact Box” next PET imaging is thus considered a “biomarker” page”). of disease in the current recommendations on diagnostic guidelines [4].

72 Chapter 6 Imaging in Neurological and Vascular Brain Diseases (PET/CT)

FACT BOX Tips and tricks for the technologist: Optimising scanner use A 10-min 18F-FDG PET brain scan can be performed 40 min p.i. Whole-body 18F-FDG PET/CT is performed 60 min p.i. Thus, brain PET/CT scanning need not occupy a time slot for whole-body PET/CT if it is performed as the fi rst scan in the morning. The two patients for brain and whole-body PET scanning are injected at the same time point with 18F-FDG when it becomes available, and the brain scanning will be fi nished before the whole-body PET scanning needs to commence. Head movements EANM The most important duty of the technologist in order to accomplish a successful and diag- nostic PET brain scan is to ensure secure head fi xation during the scanning period, and to identify any head movements that nevertheless occur. Patients with neurological diseases may have diffi culties in understanding and retaining instructions, have seizures during scanning, be agitated or suff er age-related degeneration in the spine. All of these factors may contribute to head movements during scanning. It is important (a) to identify these “at risk” patients prior to scanning through direct contact with patients, caregivers and refer- ring clinicians and inspection of patient history and (b) to take appropriate measures. Such measures might be: • Secure head fi xation in a sturdy head holder, and placement of a leg rest under the knees to prevent downward patient movement. • Placement of marks on the skin using a marking pen. • Performance of a list mode acquisition or a range of short dynamic scans, e.g. 5 × 2 min. If head movements occur during the scan, a reconstruction of the fi rst 5 min may be of suffi cient clinical quality. • Control for signifi cant movement (>5–10 mm) at the end of scanning. If such movement has occurred, repeat CT for attenuation correction alone or in combination with a new brain 18F-FDG scanning sequence depending on circumstances. If head movements occur, the CT attenuation and PET emission scans will no longer be aligned and signifi cant artefacts may be present. If these artefacts were to pass unnoticed, an errone- ous diagnosis would be made (e.g. Fig. 1). For more details, see the EANM guidelines [7].

73 FDG PET & CT-AC FDG PET Statistical Surface Projections A

B

C

Figure 1 A–C: Examples of typical 18F-FDG PET images in healthy young subjects who moved their heads after low-dose CT (CT-AC), causing misalignment and quantitative reduction in regional activity. To the right are statistical surface projections (Scenium, Siemens) comparing the subjects with a database of age-matched controls. The observed changes could have led to misdiagnosis in a clinical setting if their cause had gone unnoticed, underlining the importance of head fixation. (A) Coronal images showing the consequence of movement (1.0 cm) in the scanner in the axial direction, with false reductions in uptake in the frontal areas bilaterally (red arrows). (B) Sagittal images showing the consequence of rotation of the head anteriorly (1.2 cm), with false reductions in uptake in the mesial frontal areas bilaterally (red arrows). (C) Transverse images showing the consequence of rotation of the head to the left (1.0 cm), with false reductions in uptake in the left cortical hemisphere and false increases in uptake in the right cortical hemisphere (red arrows)

74 Chapter 6 Imaging in Neurological and Vascular Brain Diseases (PET/CT) EANM

Figure 2: Example of a typical 18F-FDG PET image in Alzheimer’s disease (AD): the posterior cingulate cortex, precuneus and lateral parietal cortex have a significantly reduced metabolism (upper row) compared with a normal reference population (lower row: the green overlay indicates areas with significant hypometabolism, as analysed by BRASS™, Hermes Medical Solutions)

PET imaging of brain glucose metabolism, of frontotemporal dementia (bvFTD), char- using 18F-fluorodeoxyglucose (18F-FDG) as acterised by hypometabolism in frontal re- the tracer, is a validated and routinely used gions, and corticobasal degeneration (CBD), method to show regional abnormalities, which typically shows asymmetrical cortical some of which are typical for the different de- (frontoparietal) and subcortical hypometab- mentia syndromes [5–7]. AD is characterised olism contralateral to the affected hemibody by hypometabolism involving the temporo- (Fig. 3). The visual identification of these spe- parietal association cortices, in particular the cific patterns can be supported by various precuneus and posterior cingulate cortex (an automated methods, usually based on the example is provided in Fig. 2). Patients with normalisation of individual images to a refer- DLB have a pattern similar to AD patients, the ence space and then on the semi-quantifica- hypometabolism often extending to the oc- tion of abnormalities relative to the distribu- cipital cortex and possibly sparing the pos- tion observed in a group of healthy subjects terior cingulate region. FTLD includes various [8, 9]. An example is provided in Fig. 2. syndromes, such as the behavioural variant

75 Figure 3: Example of the hypometabolism observed in cases of corticobasal degeneration (CBD) involving the frontoparietal cortex and the basal ganglia of the left hemisphere (arrows)

During the last decade, specific tracers able and patients with AD and to identify amyloid to visualise amyloid plaques in vivo have deposits in vivo, as compared with results at been tested and validated in humans. The autopsy [13, 14]. Different strategies for visual first was the 11C- (11C- reading of the images have been proposed PIB), which was able to bind fibrillary amyloid for the different tracers, based on the use with a good correlation with post-mortem of quantitative indices such as the ratio of measures [10–12]. More recently, 18F-labelled the mean standardised uptake value (SUV) compounds, patented by industrial com- across cortical regions to the SUV measured panies, have been approved by regulatory in a reference region, typically the cerebel- agencies both in Europe and in the United lum. In general, a positive image shows sig- States for use in patients with prodromal or nificant cortical uptake of the tracer, while atypical dementia. These tracers have been a negative image shows variable uptake in validated in phase II and phase III trials for white matter and no relevant cortical signal their ability to discriminate healthy controls (examples are provided in Fig. 4).

76 Chapter 6 Imaging in Neurological and Vascular Brain Diseases (PET/CT) EANM

Figure 4: Examples of typical positive (upper row) and negative (lower row) amyloid PET images, obtained using an 18F-labelled tracer [in this case florbetapir (AV-45)]

Amyloid imaging is considered an early be at higher risk for subsequent development marker of the AD pathological process, reach- of dementia, but this is still a debated topic, ing the threshold of positivity about 17 years currently under investigation. First, the expres- before the clinical onset of overt dementia sion of clinical disease is due to not only the [15]. Appropriate criteria for the clinical use amount of “pathology” in the brain but also of amyloid PET imaging have been published the capacity of the brain to cope with dam- recently [16]. In particular, this examination is age, called the “reserve capacity”. The concept advised for patients with persistent or progres- of cognitive reserve is based on the observa- sive cognitive impairment that has an atypical tion that individuals with a high educational or unclear clinical presentation or early onset level and intelligence preserve a normal (before 65 years of age). In subjects with MCI, functional level for longer than less educated amyloid imaging has a high positive predictive people, despite neurodegeneration [19]. Vari- value for progression to AD [12]. A consistent ous PET studies have shown that cognitive finding across different amyloid PET imaging reserve modulates the interaction between studies is that around 30% of cognitively nor- some measures of pathology and clinical se- mal elderly subjects have a positive scan, in verity [20–22]. Second, amyloid deposits are agreement with the proportion of cognitively possibly the first event in the pathological cas- normal elderly subjects with an autopsy diag- cade of AD, but by themselves do not account nosis of AD [17, 18]. These individuals might for cognitive decline: other factors, such as the

77 aggregation of hyperphosphorylated tau pro- ing modalities of choice for this purpose. The tein, play a central role in the onset and pro- intravenous administration of the tracer has gression of neurodegeneration [2]. New PET to be performed during EEG monitoring of tracers aimed at imaging tau aggregates are the epileptic activity, which is a major deter- currently under development and preliminary minant of the imaging findings. Indeed, while results suggest that they may provide crucial an “ictal” injection, during a seizure, will show information about the interplay of amyloid the increased perfusion and metabolism in deposits and the earliest clinical signs [23]. the epileptic focus, an “interictal” injection will depict the dysfunctional cortex (often includ- Finally, a large panel of other PET tracers rele- ing the epileptic focus and some regions of vant to the investigation of dementias exists, seizure propagation) as a hypometabolic area e.g. tracers able to visualise different neuro- compared with the surrounding cortex [26]. transmission systems, such as the cholinergic Interictal imaging is preferable, given that truly or dopaminergic system, or to measure the ictal images require coordination of radiotrac- occurrence of neuroinflammation, which is er availability and seizure onset and duration, presumably a factor contributing to the pro- which is difficult to assure. Interictal 18F-FDG gression of the neurodegenerative process PET has a sensitivity above 80% for identifica- [24]. These tracers, which are of utmost inter- tion of the epileptic focus in epi- est for a better understanding of the patho- lepsy [27–30]. The contribution of 18F-FDG PET logical processes of dementia, are still limited imaging is particularly relevant when no lesion to dedicated research centres and are being is identified on magnetic resonance imaging evaluated in clinical trials. (MRI) or in the case of multiple abnormalities as it limits the need for invasive recording [26, 31]. Epilepsy Indeed, it has been reported that MRI-negative Epilepsy is one of the most common chronic temporal lobe epilepsy patients with clear neurological conditions, affecting 0.7% (0.5– unilateral anterior temporal hypometabolism 1%) of the population. It has very significant show the same positive outcome as patients social and professional consequences and is with a morphologically visible lesion [32]. An associated with increased mortality. Around example of a “non-lesional” patient with a posi- 30–40% of patients with epilepsy suffer from tive PET is shown in Fig. 5. PET can also be deci- focal seizures that are refractory to anti-epi- sive in situations of multifocal lesions or multi- leptic drug treatment [25]. In these patients, focal epileptic activity. In tuberous sclerosis, an surgical resection of the epileptic focus is the autosomal dominant disorder characterised only treatment that can possibly cure the con- by multiple cortical malformations (tubers), dition, and its success depends strongly on the most hypometabolic tuber is concordant accurate presurgical localisation of the focal with invasive localisation of the epileptic focus abnormality. 18F-FDG PET is one of the imag- in a majority of patients [33].

78 Chapter 6 Imaging in Neurological and Vascular Brain Diseases (PET/CT) EANM

Figure 5: Example of a patient with no visible lesion on MRI (upper row) but with clear hypometabolism in the left temporal pole on 18F-FDG PET (indicated by arrows, lower row), concordant with the clinical and EEG findings

Other PET tracers have shown promising Amyotrophic lateral sclerosis results in specific applications: for example, Amyotrophic lateral sclerosis (ALS) is a neuro- 11C-flumazenil, a GABA antagonist, could be degenerative disease that has an incidence useful to localise pathological brain tissue in of 2–3/100,000 new cases per year [35] and is temporal lobe epilepsy, with increased up- characterised by progressive muscle paralysis take on the pathological side [34]. 11C-AMT associated with degeneration of motor neu- (α-[¹¹C]-methyl-L-) displays posi- rons in primary motor cortex, brainstem and tive uptake in the most epileptogenic tuber spinal cord [36]. ALS-mimicking conditions in tuberous sclerosis in about 70% of cases include pure upper motor neuron disease [33]. These tracers are used in specific re- (primary lateral sclerosis, PLS), pure lower search settings and are not currently applied motor neuron disease (progressive muscu- in clinical practice. lar atrophy) and FTLD, the latter sharing with ALS similar cognitive symptomatology.

79 PET represents a valuable tool in the diagnos- A tic work-up and for the differentiation of ALS from mimicking diseases. Various radiotrac- ers (11C-flumazenil [37],11 C- (R)PK11195 [38], 18F-DPA-714 [39], 11C-(L)-deprenyl-D2 [40]) have been used to assess changes in regions affected by ALS, including those for the eval- uation of activity in the white matter [38–40], which plays a significant role in the progres- sion of the disease. However, most of the PET studies in ALS have been performed using 18F-FDG, focussing on patients and normal controls as a comparison group and looking B for ALS-related metabolic patterns. Patients with ALS usually demonstrate diffusely de- creased 18F-FDG uptake in primary and sup- plementary motor and premotor cortices as well as clusters of relative hypermetabolism in the mesiotemporal regions, occipital cor- tex, cerebellum, corticospinal tracts and up- per brain stem (Fig. 6), reflecting a complex neuropathophysiology involving degenera- tion of grey matter and areas of reactive mi- croglial activation. The presence of astrocy- C tosis and activated microglia in corticospinal 5 tract, midbrain and pons has also been sup- 1 ported by findings from H-magnetic reso- 4 nance spectroscopy (increased myo-inositol [41]) and other PET investigations (increased 3 uptake in peripheral benzodiazepine recep- tors [38, 39] and astrocytes [40]). 2

1

0 Figure 6: Comparison of 18F-FDG PET findings in ALS patients and controls using statistical

80 Chapter 6 Imaging in Neurological and Vascular Brain Diseases (PET/CT)

parametric mapping. Statistically significant In addition, PET has a promising role in dis- differences (p<0.05 false discovery rate criminating ALS from ALS-mimicking pathol- corrected) are highlighted on an MRI T1 ogies, thereby contributing to recruitment template. (A) sagittal view; (B) coronal view; of suitable subjects for clinical trials and al- (C) transverse view. Clusters of significant lowing early interventions when appropriate hypermetabolism are seen in the midbrain, [45]. Longitudinal examinations, identifying cerebellum, medial temporal lobes and significant changes over time, and novel corticospinal tracts PET/MRI technology, providing a simultane- Recently the potential of 18F-FDG as an im- ous assessment of structural and functional portant biomarker for the diagnosis of ALS lesions, will in the future play a pivotal role in has been confirmed by two investigations in better understanding the pathophysiology which extremely large patient cohorts were of this complex and fatal disease. EANM recruited [42, 43]. Both studies analysed the diagnostic ability of PET in ALS and found Stroke that it showed a sensitivity of about 95% in Background separating patients from controls [42, 43] Stroke is the third leading cause of death and of 71% in separating ALS from PLS [43]. globally and the second most frequent cause Pagani et al. [42] reported in 195 ALS patients of death in the developed world after coro- a mixed 18F-FDG hypermetabolic–hypometa- nary artery disease [46]. It is also a major cause bolic pattern, with marked hypometabolism of chronic disability, particularly among the in the frontal, premotor and occipital cortices elderly population. The annual stroke rate is and relative hypermetabolism in white mat- in the range of 1.4–4 per 1000. Approximate- ter, midbrain and the medial temporal cortex. ly 30% of cases are attributed to cerebral haemorrhage and 70% to cerebral ischaemia. PET findings have also been compared in pa- Stroke may arise as a result of a combination tients with sporadic ALS without mutations of different factors. Ischaemic strokes can of ALS-related genes, patients with ALS and be produced through thrombotic, embolic comorbid FTD and patients with ALS carry- and haemodynamic mechanisms. Clinically ing the C9orf72 mutation [44]. More wide- infarcts are commonly considered as athero- spread cortical and subcortical involvement, thrombotic (local thrombosis in relation to especially in the frontotemporal cortex, was atheroma of the vascular wall), cardioembolic associated with a more severe clinical course (embolism of cardiac origin, e.g. atrial fibrilla- in those with the C9orf72 mutation. tion, recent , aortic valve

81 disease) or lacunar (occlusion of one of the Nuclear imaging has been used extensively small penetrating end-arteries at the base in the study of neurovascular diseases over of the brain, often resulting from microath- the past five decades and has yielded impor- eroma formation). A lesion between 3 and 15 tant knowledge on the pathophysiology of mm in diameter is commonly regarded as a acute stroke. In clinical practice, however, the lacune. Occasionally, an ischaemic infarction primary methods of choice in the diagnostic may turn into a haemorrhage. The clinical work-up of acute stroke and neurovascular symptoms are of sudden onset and depend diseases are CT and MRI scanning. At the mo- on the location and size of the lesion. Infarcts, ment non-enhanced CT of the brain remains particularly the lacunar ones, may be clinical- the mainstay of imaging in the setting of an ly silent or transient. If the symptoms last <24 acute stroke (Fig. 7) since it is fast, inexpensive h, the term transient ischaemic attack (TIA) and readily available. Its main limitation is the is used. While TIAs generally do not cause low sensitivity in the acute setting, where MRI permanent brain damage, they are a seri- has an advantage. PET/CT is not indicated in ous warning sign and should not be ignored. acute stroke. However, diagnostic CT of the There are many risk factors for stroke: age, brain is easily available in all PET/CT systems gender, ethnicity, heredity, hypertension, and can provide important additional infor- cigarette smoking, hyperlipidaemia, diabetes mation relevant to the interpretation of any mellitus, obesity, fibrinogen and clotting fac- functional defects found, e.g. in dementia. tors, oral contraceptives, erythrocytosis and Knowledge of the various presentations of haematocrit level, prior cerebrovascular and neurovascular disease becomes important in other diseases, physical inactivity, diet and the differential diagnosis of vascular demen- alcohol consumption, illicit drug use and ge- tia (VaD), AD and mixed dementia (AD with netic predisposition [47]. concomitant stroke or small vessel disease).

82 Chapter 6 Imaging in Neurological and Vascular Brain Diseases (PET/CT)

T2 MRI RCBF Baseline RCBF Acetazolamide CT 2 mth post PET EANM

R L

Figure 7: Patient with diabetes mellitus, hypertension and multiple previous TIAs. CT angiography showed occlusions of the right internal carotid and right vertebral arteries. There are typical watershed infarctions in the centrum semiovale on T2-weighted MRI (red arrows). PET measurements of rCBF at baseline using 15O-water show discrete hypoperfusion in the right hemisphere, which is particularly pronounced above the infarcts (white arrow). During acetazolamide stimulation, the perfusion was increased by 60% in the left hemisphere, but decreased by 9% in the right middle cerebral artery territory, consistent with cerebrovascular steal (green arrow, same scale as baseline). The patient refused reconstructive vascular surgery and sustained an ischaemic stroke during an -induced hypotensive episode in the same risk areas as had been defined on PET 2 months earlier (hypointense lesions, orange arrows), demonstrating the clinical validity of the method

Furthermore, quantitative measurements of nuclear imaging for evaluation of the hae- the regional cerebral blood flow (rCBF) us- modynamic response in patients with cere- ing positron emission tomography (PET) still brovascular occlusive disease prior to revas- occupy a specialised clinical niche within cularising surgery.

83 Dementia and stroke forms are crossed cerebro-cerebellar diaschisis Many of the risk factors mentioned in the [53], where damage to one hemisphere leads preceding section are shared between de- to depression of the contralateral cerebel- mentia and stroke. About 30% of dementia lar hemisphere; thalamo-cortical diaschisis, patients will show signs of cerebrovascular where damage/stroke in the thalamus leads disease that may be synergistic with AD in to depression of the ipsilateral hemisphere producing the clinical syndrome of demen- [54]; and cross cerebello-cerebral diaschisis, tia. Mixed dementia is the second most com- where damage to one cerebellar hemi- mon form of dementia after AD (10–20%), sphere leads to depression of the contralat- while VaD alone accounts for 10% of cases. eral cerebral hemisphere [55]. Furthermore, VaD increases in prevalence with age [48]. It decreased activity in cortical areas may be may arise from multiple cortical infarctions, evident if the infarct is located subcortically, strategic strokes and subcortical white mat- involving and undercutting white matter ter lesions [49]. The neuropsychological char- tracts (Fig. 8). Thus, for each individual area of acteristics of VaD may be different from those decreased metabolic activity in patients eval- seen in AD. Disturbances in frontal-executive uated for dementia using 18F-FDG PET/CT, it functions (Fig. 8), rather than memory, are of- is necessary to determine whether the re- ten the more dominant feature and memory duction in activity can be explained by local impairment may be absent in some patients or distant neuronal damage/vascular lesion. with significant cognitive deficits.

Vascular disease will often give rise to a re- gional decrease in activity encompassing the infarcted area and neighbouring areas to varying degrees (Fig. 8). Cases of vascular disease may also be metabolically silent. It is important to be familiar with the concept of diaschisis, the idea that damage to one part of the nervous system can have effects at a distance due to loss of input [50]. Thus, it was demonstrated as long ago as 1964, using the Lassen-Ingvar krypton-85 method [51], that stroke patients had a strikingly low rCBF measurement also in the structurally intact healthy hemisphere [52]. The most common

84 Chapter 6 Imaging in Neurological and Vascular Brain Diseases (PET/CT)

FDG PET & T1 MR Statistical Surface Projections A

B CT FDG PET EANM

C

Figure 8 A–C: Examples of typical 18F-FDG PET images in patients referred for PET/CT or PET/MRI for evaluation of dementia. To the right are statistical surface projections (Scenium, Siemens) comparing the subjects with a database of age-matched controls. (A) PET fusion with simultaneously acquired T1-weighted MRI: transverse images. The red arrows indicate circumscribed lacunar infarcts with decreased activity uptake involving the left head of the caudate and the anterior thalamus; these lesions are undercutting projections to the structurally intact left frontal lobe, giving rise to a moderate reduction in activity (green arrows). FTD could have been considered, but this is vascular dementia. There is also an infarct in the pons (not shown). (B) CT and PET transverse images. CT shows extensive subcortical hypointense signals (leukoaraiosis) indicative of subcortical ischaemia (red arrows) that cannot alone explain the pattern of cortical metabolic reduction (green arrow), suggesting a mixed vascular and neurodegenerative origin. (C) CT and PET/CT transverse images show a subcortical infarct in the right temporal region (red arrow), leading to a larger metabolic defect in the temporoparietal cortex (green arrow) through disruption of subcortical–cortical circuits. The PET image alone could be misread as demonstrating signs of neurodegeneration in the absence of supportive correlation with CT

85 Chronic cerebrovascular disease static limb shaking, impaired vasoreactivity to Atherosclerotic internal carotid artery occlu- acetazolamide challenge and increased rOEF sion causes approximately 10% of TIAs and on PET scanning [60]. 15–25% of ischaemic strokes in the carotid territory. The 2-year risk of ipsilateral ischaemic In the event of carotid occlusion, the affected stroke while a patient is receiving medical hemisphere receives its blood supply from therapy is 10–15% [56]. This risk, however, de- communicating arteries of the circle of Willis pends on the capacity of the brain tissue to and collaterals between the extracranial and compensate, e.g. by increasing the regional intracranial arteries. If these are not sufficiently oxygen extraction fraction (rOEF). Increased developed, there will be a pressure drop in the rOEF indicates that the last defence before most distant arterioles of the arterial tree at the stroke has been mobilised. In symptomatic border zones between vascular territories (wa- patients with increased rOEF, the 2-year risk of tershed areas). This will initially be compensat- ipsilateral ischaemic stroke is 30–40%, while it ed haemodynamically by vasodilation, which is only 5% in symptomatic patients with nor- will stabilise rCBF and can be measured as an mal rOEF [57]. In the event of symptomatic increase in regional cerebral blood volume occlusions of the carotid arteries to the brain, (Fig. 9). Metabolic compensation will follow a revascularising surgical procedure may be via an increase in oxygen extraction from the considered: the so-called extracranial–intra- capillary blood, the rOEF rising from 30–40% cranial bypass operation (EC-IC). In the most to 70–80%. In patients with exhausted perfu- common version of the EC-IC bypass opera- sion reserve, rCBF will behave in a pressure- tion, a branch of the external carotid artery, passive manner. The arterioles are maximally usually the superficial temporal artery, is dilated, and rCBF will change with the perfu- anastomosed to a branch of the internal ca- sion pressure. If the patient experiences a lon- rotid artery on the occluded side, usually the ger lasting pressure drop (e.g. due to systemic meningeal artery. The procedure is not with- infection, cardiovascular disease, dehydration out risk, and there is a 10–15% likelihood of or blood pressure-lowering drugs) and the perioperative stroke [58, 59]. Thus, the overall compensatory mechanisms are insufficient, risk is the same as in medically treated symp- an ischaemic infarction will develop (Fig. 7) tomatic patients over 2 years. Consequently, [61, 62]. Acetazolamide (Diamox) is a carbo- only the subgroup of high-risk patients should anhydrase inhibitor and a potent vasodilator be considered for the operation. Supportive of the cerebral vessels that increases rCBF by symptoms and findings that may predict hae- 20–60%. Only arterioles that are not already modynamic failure and high risk of stroke and dilated will respond to acetazolamide, while death are usually manifested as repeated TIAs the in affected re- or stroke, with typical subcortical “watershed” gions will be either reduced, unaffected or infarcts on T2-weighted MRI (Fig. 7), ortho- negative. The last-mentioned phenomenon

86 Chapter 6 Imaging in Neurological and Vascular Brain Diseases (PET/CT)

is called “cerebrovascular steal” or the “reverse PET measurements of increased rOEF in the Robin Hood effect”. The increased rCBF in the symptomatic hemisphere (the COSS study). unaffected hemisphere decreases perfusion The study endpoint was 2-year stroke recur- pressure and rCBF in the affected hemisphere rence, the rate of which was found to be 21% (Fig. 9). EC-IC bypass surgery can reverse these in the surgical group and 23% in the medically changes [63] and may increase the cognitive treated group. The latter was far below the performance [64], but the clinical benefit in expected rate of 40% from prospective ob- terms of stroke risk and survival is still con- servational investigations, and the study failed troversial. In a recent randomised clinical trial, to demonstrate a significant difference in the 195 patients were randomised to either opti- two treatments [58]. mal medical treatment or surgery based on EANM

Figure 9: A schematic representation of the vascular response to carotid occlusion and vasodilation using acetazolamide. The two internal carotid arteries are represented as two arrows connected by the circle of Willis, which subsequently supplies feeding arteries to the two hemispheres. The bars in the hemispheres represent the vasodilatory state of the arterioles in brain tissue. In the healthy brain (top row), acetazolamide dilates the arterioles and increases rCBF symmetrically in

87 the two hemispheres by 20–60%. In carotid artery occlusion (lower row) – represented by the black box in the left carotid – the ipsilateral hemisphere is fed through a more or less patent collateral blood supply in the circle of Willis; these arterioles are dilated because of the ensuing decrease in perfusion pressure, which stabilises rCBF. When the arterioles in the unaffected contralateral hemisphere are dilated, the perfusion pressure will decrease even further ipsilaterally. Depending on the residual vasodilatory capacity, the rCBF will be increased, unaffected or decreased. The last- mentioned phenomenon is known as the cerebrovascular steal phenomenon

The measurement of vasoreactivity in re- mised haemodynamic reactivity is often in sponse to acetazolamide requires quanti- the watershed areas, the subtraction image tative assessment of the rCBF using either is the most efficient way of directly revealing 15O-water PET or the stable xenon-CT SPECT the location of increased and decreased rCBF. method [65]. Both methods are available only in specialised units. 15O-water has a half- Acetazolamide challenge is also used in com- life of only 2 min and an on-site is bination with transcranial Doppler flow mea- required for on-line continuous production surements. To date there have been virtually of the tracer. Furthermore, quantification of no reports of adverse effects of acetazolamide rCBF measured in mL blood flow per 100 g challenge [66]. One case report did, however, tissue per minute necessitates arterial cannu- cite a stroke episode 24 h after injection, lation and continuous blood sampling with which could have been related to the mildly radioactivity measurements for kinetic mod- diuretic effect of acetazolamide [67]. elling with calculation of parametric images.

In the acetazolamide challenge, rCBF is mea- sured at rest and approximately 20 min after injection of 500–1500 mg acetazolamide i.v. The baseline rCBF image is subtracted from the acetazolamide-stimulated rCBF image, and all images are fused to T2-weighted/ FLAIR MRI. The baseline rCBF image allows estimation of the extent of infarction and hypoperfusion around the areas of signal change on MRI. Regions of interest are drawn on the three major intracerebral artery terri- tories in both hemispheres for quantification of rCBF and reactivity. However, as compro-

88 10. 9. 6. 5. 4. 2. 1. References References Chapter 6 3. 11. 8. 7.

McKhann GM, Knopman DS, Chertkow H, Hyman BT, H,Hyman DS,Chertkow GM,Knopman McKhann WHO. Fact Dementia, sheetN°362.2012. Klunk Klunk WE, EnglerH,Nordberg A, Wang Y, Blomqvist 2013: whatradiologists needto know. Eur Radiol. C,etal. NeuroimagingA, Rodriguez ofdementiain Haller S,Garibotto V, C,Xekardaki E,Bouras Kovari Lucignani G,NobiliF. for assessment early FDG-PET CRJr,Jack DS,Jagust Knopman WJ, Petersen RC, Leifer BP. diagnosis Early ofAlzheimer’s disease:clinical 2013;12:207–16. Lancet Neurol.thetical model of dynamic biomarkers. cal processes in Alzheimer’s disease: an updated hypo Weiner MW, AisenPS,etal. Tracking pathophysiologi- drug development. Alzheimers Res drug development. AlzheimersRes Ther. 2011;3:34. for diagnosis early and onbiomarkers perspectives new Nordberg imaging A.Molecular inAlzheimer’s disease: 2004;55:306–19. disease withPittsburgh Compound-B. AnnNeurol. G, HoltDP, etal. brainamyloid Imaging inAlzheimer’s setting. Neuroimage Clin.2014;6:445–54. cedure indementia diagnosis for in a clinical FDG-PET F, Vanoli EG, et al. Validation of an optimized SPM pro zheimers Dis. 2015;44:815-26. project.JAl Alzheimer’s (EADC) DiseaseConsortium inamnesticMCI:anEuropean analysis of18F-FDG-PET GB, S,etal. Galluzzi Visual versus semi-quantitative I,Buschiazzo A,Frisoni S,Brugnolo A,Bossert Morbelli Imaging. 2009;36:2103–10. Mol brain imaging using [18F]FDG, version 2. Eur J Nucl Med F, etal. Nagren EANMprocedure K, guidelinesfor PET Varrone A, Asenbaum S, Vander Borght T, J, Booij Nobili 2013;23:3393–404. Imaging. 2010;37:1604–9. Mol enough? Eur JNuclMed of Alzheimer’s disease:isn’t baselarge theevidence disease. 2011;7:263–9. AlzheimersDement. workgroups ondiagnostic guidelinesfor Alzheimer’s onAging-Alzheimer’sNational Institute Association due to Alzheimer’s disease:recommendations from the CRJr,Jack CH,etal. Kawas The diagnosis ofdementia Suppl Dementia):S281–8. Soc. and economicbenefits. 2003;51(5 Geriatr JAm Perani D, PA, Rosa Della F, Cerami C,Gallivanone Fallanca - - - 89 17. 16. 15. 12. 21. 19. 18. 14. 13. 20.

Villemagne S,Bourgeat P,VL, Burnham Brown B, Ellis KA, Johnson KA, Minoshima S, Bohnen NI, Donohoe KJ, KJ, NI,Donohoe S,Bohnen Minoshima KA, Johnson Stern Stern Y. Cognitive Neuropsychologia. reserve. Tsopelas ND, etal. Frequent amyloid depositionwithout 2013;9:e-1–16. Dement. Imaging, and the Alzheimer’s Association. Alzheimers Force, andMolecular ofNuclearMedicine theSociety for oftheAmyloid amyloid Imaging areport PET: Task Foster NL, Herscovitch P, etal. Appropriate usecriteria 2013;12:357–67. study.disease: aprospective cohort Lancet Neurol. eration, andcognitive declineinsporadicAlzheimer’s Salvado O, etal. Amyloid betadeposition,neurodegen phase 2trial. AnnNeurol. 2010;68:319–29. in Alzheimerdiseaseandmildcognitive a impairment: tin C, Triau E,etal. 18F-flutemetamol amyloid imaging Vandenberghe R, Van A, Salmon E, Bas- Ivanoiu Laere K, beta-amyloid pathology. JAMA. 2011;305:275–83. MA, et al.Mintun for imaging Useof florbetapir-PET BJ, JA,Bedell CM,Schneider Beach Clark TG, Bilker WB, Imaging. 2013;40:104–14. Mol Nucl Med amyloidstudy of fibrillar in Alzheimer’s disease. Eur J Vandenberghe R,etal. AEuropean multicentre PET Neurology. 2008;71:1342–9. evidence. and AD: FDG-PET in aMCI converters reserve Holtoff V, etal. Education andoccupation asproxies for Med. 2013;54:894–902. project.JNucl European Alzheimer diseaseconsortium cognitive inprodromal reserve Alzheimerdisease:a underlying BN,etal.van Berckel networks Metabolic A,Frisoni S,PerneczkyR,Drzezga Morbelli GB, Caroli A, 2009;47:2015–28. 2011;134(Pt 3):798–807. phase of Alzheimer’s disease. Brain. the pre-dementia declinein poral beta-amyloid depositionto memory P, oftem contribution G,etal. Jones Independent Arch Neurol. 2008;65:1509–17. significant cognitive amongtheelderly. impairment Nordberg SF, A, Carter J, Rinne A,Brooks Drzezga DJ, Chetelat G, Villemagne VL, Pike Bourgeat KE,Ellis KA, Aizenstein HJ, NebesRD, JA,Price CA, Saxton JC,Mathis Garibotto Garibotto V, Salmon E, K, E, Herholz Borroni B, Kalbe - - EANM 22. Garibotto V, Tettamanti M, Marcone A, Florea I, Panzacchi 34. Hammers A, Koepp MJ, Hurlemann R, Thom M, Richard- A, Moresco R, et al. Cholinergic activity correlates with son MP, Brooks DJ, et al. Abnormalities of grey and white reserve proxies in Alzheimer’s disease. Neurobiol Aging. matter [11C]flumazenil binding in temporal lobe epi- 2013;34:2694.e13–8. lepsy with normal MRI. Brain. 2002;125(Pt 10):2257–71.

23. Zwan MD, Okamura N, Fodero-Tavoletti MT, Furumoto 35. Chio A, Mora G, Calvo A, Mazzini L, Bottacchi E, Mutani R, S, Masters CL, Rowe CC, et al. Voyage au bout de la nuit: et al. Epidemiology of ALS in Italy: a 10-year prospective Abeta and tau imaging in dementias. Q J Nucl Med Mol population-based study. Neurology. 2009;72:725–31. Imaging. 2014;58:398–412. 36. Brooks BR, Miller RG, Swash M, Munsat TL, World Federa- 24. Baskin A, Giannakopoulos P, Ratib O, Seimbille Y, Assal F, tion of Neurology Research Group on Motor Neuron D. Perani D, et al. PET radiotracers for molecular imaging El Escorial revisited: revised criteria for the diagnosis of in dementia. Curr Radiopharm. 2013;6:215–30. amyotrophic lateral sclerosis. Amyotroph Lateral Scler Other Motor Neuron Disord. 2000;1:293–9. 25. Kwan P, Brodie MJ. Early identification of refractory epi- lepsy. N Engl J Med. 2000;342:314–9. 37. Lloyd CM, Richardson MP, Brooks DJ, Al-Chalabi A, Leigh PN. Extramotor involvement in ALS: PET studies with the 26. Kumar A, Semah F, Chugani HT, Theodore WH. Epilepsy GABA(A) ligand [(11)C]flumazenil. Brain. 2000;123:2289– diagnosis: positron emission tomography. Handb Clin 96. Neurol. 2012;107:409–24. 38. Turner MR, Cagnin A, Turkheimer FE, Miller CC, Shaw 27. Ryvlin P, Bouvard S, Le Bars D, De Lamerie G, Gregoire CE, Brooks DJ, et al. Evidence of widespread cerebral MC, Kahane P, et al. Clinical utility of flumazenil-PET ver- microglial activation in amyotrophic lateral sclerosis: an sus [18F]fluorodeoxyglucose-PET and MRI in refractory [11C](R)-PK11195 positron emission tomography study. partial epilepsy. A prospective study in 100 patients. Neurobiol Dis. 2004;15:601–9. Brain. 1998;121 (Pt 11):2067–81. 39. Corcia P, Tauber C, Vercoullie J, Arlicot N, Prunier C, Pra- 28. Knowlton RC. The role of FDG-PET, ictal SPECT, and line J, et al. Molecular imaging of microglial activation in MEG in the epilepsy surgery evaluation. Epilepsy Behav. amyotrophic lateral sclerosis. PLoS One. 2012;7:e52941. 2006;8:91–101. 40. Johansson A, Engler H, Blomquist G, Scott B, Wall A, 29. Henry TR, Engel J, Jr., Mazziotta JC. Clinical evaluation of Aquilonius SM, et al. Evidence for astrocytosis in ALS interictal fluorine-18-fluorodeoxyglucose PET in partial demonstrated by [11C](L)-deprenyl-D2 PET. J Neurol epilepsy. J Nucl Med. 1993;34:1892–8. Sci. 2007;255):17–22.

30. Kurian M, Spinelli L, Delavelle J, Willi JP, Velazquez M, 41. Kalra S, Hanstock CC, Martin WR, Allen PS, Johnston Chaves V, et al. Multimodality imaging for focus localiza- WS. Detection of cerebral degeneration in amyotrophic tion in pediatric pharmacoresistant epilepsy. Epileptic lateral sclerosis using high-field magnetic resonance Disord. 2007;9:20–31. spectroscopy. Arch Neurol. 2006;63:1144–8.

31. Chassoux F, Rodrigo S, Semah F, Beuvon F, Landre E, 42. Pagani M, Chio A, Valentini MC, Oberg J, Nobili F, Calvo Devaux B, et al. FDG-PET improves surgical outcome A, et al. Functional pattern of brain FDG-PET in amyo- in negative MRI Taylor-type focal cortical dysplasias. trophic lateral sclerosis. Neurology. 2014;83:1067–74. Neurology. 2010;75:2168–75. 43. Van Laere K, Vanhee A, Verschueren J, De Coster L, 32. Carne RP, O’Brien TJ, Kilpatrick CJ, MacGregor LR, Hicks Driesen A, Dupont P, et al. Value of 18fluorodeoxyglu- RJ, Murphy MA, et al. MRI-negative PET-positive tem- cose-positron-emission tomography in amyotrophic poral lobe epilepsy: a distinct surgically remediable lateral sclerosis: a prospective study. JAMA Neurol. syndrome. Brain. 2004;127(Pt 10):2276–85. 2014;71:553–61.

33. Juhasz C, Chugani DC, Muzik O, Shah A, Asano E, Mang- 44. Cistaro A, Pagani M, Montuschi A, Calvo A, Moglia C, ner TJ, et al. Alpha-methyl-L-tryptophan PET detects epi- Canosa A, et al. The metabolic signature of C9ORF72- leptogenic cortex in children with intractable epilepsy. related ALS: FDG PET comparison with nonmutated Neurology. 2003;60:960–8. patients. Eur J Nucl Med Mol Imaging. 2014;41:844–52.

90 57. 56. 55. 54. 53. 49. 48. 46. 45. References Chapter 6 47. 52. 51. 50.

Sachdev P,Sachdev J, I,AlladiS,Black Skoog R,O’Brien Kalaria Feeney DM,Baron JC.Diaschisis. Stroke. 1986;17:817–30. Bousser MG.Effects ofthalamicstroke onenergyBousser me Baron JC, D’Antona R, Pantano P, Serdaru M, Samson Y, 2014;13:1228–40. changes. andfunctional LancetNeurol.into structural Neuroimaging inamyotrophic lateral sclerosis: insights Chio A,Pagani M, Agosta F, Calvo A,Cistaro A, Filippi M. Powers WJ. occlusion. Curr Atherosclerotic carotid artery Desmond DW.Desmond Vascular inevolu- dementia:aconstruct Grubb RL,Jr., Derdeyn CP, Fritsch SM,Carpenter DA, Broich K, Hartmann A,BiersackHJ, R.Crossed Horn Hartmann Broich K, -Part IofIII.Angiology. 2000;51:793–808. Stroke: epidemiology, clinicalpicture, factors- andrisk Tegos AN. SS,Nicolaides TJ, E,Daskalopoulou Kalodiki of DiseaseStudy2013.Lancet.2014;385:117–71. 1990–2013: asystematic analysisfor theGlobalBurden for 240causesofdeath, mortality and cause-specific sion. JAMA. 1998;280:1055–60. intheprognosisfactors ofsymptomatic carotid occlu- Yundt KD, Videen TO, ofhemodynamic etal. Importance Treat OptionsNeurol. 2003;5:381–9. lar infarction.Neurosci Lett. 1987;83:7–12. diaschisisinapatient withcerebel- cerebello-cerebral study inman.Brain.1986;109(Pt6):1243–59. Apositrontabolism ofthecerebral tomography cortex. Neurol Scand. 1964;40:41–6. of the cerebral metabolism in man. Acta hemispheric E. Skinhoj TransneuralK, Hoedt-Rasmussen depression 1961;17:42–3. Experientia. bydetermined krypton. radioactive cortex DH. Lassen NA,Ingvar The bloodflow ofthecerebral 2004;61:283–8. of diaschisis:origins andperspectives. Arch Neurol. Finger PJ, S,Koehler C. Jagella concept The Monakow 2014;28:206–18. ders: a VASCOG statement. AlzheimerDisAssoc Disord. SE, etal. Diagnostic for criteria vascularcognitive disor tion. Cerebrovasc Rev. BrainMetab 1996;8:296–325. Global, regional, and national age-sex specific all-cause Global, regional, specificall-cause andnationalage-sex - - 91 63. 61. 58. 67. 66. 65. 64. 62. 60. 59.

D, Chang Y, for OEFreactivity hemodynamic etal. PET Nemoto EM, Yonas H,Pindzola H,Sashin RR,Kuwabara neurologic acetazolamide deficit after challenge for Choksi V, L,Hoeffner E. HughesM, Selwa Transient emission tomography-derived cerebral bloodflow M.PressureCzosnyka autoregulation andpositron P, PS,Smielewski Minhas PJ, Kirkpatrick Pickard JD, K, K, Yamadate on etal. Effects ofEC-IC bypass surgery K, Okuguchi Kuroda K, T,Sasoh M,OgasawaraK, Terasaki endarterectomy. J Vasc Surg. 1992;15:753–4. beforeation of cerebral carotid and after perfusion evalu ton emission computed tomography (SPECT) Waymire B, etal. Acetazolamide enhancedsinglepho occlusion. Lancet.1984;323:310–4. inpatientswithcarotid-artery reserve cerebral perfusion Gibbs JM, Wise RJ, Leenders KL,Jones T. Evaluation of ing. 2007;17:54–60. compromise inocclusive vasculardisease. JNeuroimag clusive disease. JStrokeCerebrovasc Dis. 2008;17:340–3. inpatientswithathero-oc hemodynamic insufficiency ML, Connolly andclinicalpredictors ES.Radiographic of MB, Otten RM,Merkow RJ, MC,Komotar Garrett Starke N EnglJMed. 1985;313:1191–200. tional randomized trial. Study Group.The EC/IC Bypass ofischemicstroke.- ofaninterna duce therisk Results Failure bypass to re arterial ofextracranial–intracranial JAMA. 2011;306:1983–92. Studyrandomized trial.the Carotid Surgery Occlusion stroke prevention inhemodynamiccerebral ischemia: Derdeyn CP. for bypass surgery Extracranial–intracranial Assist Tomogr.2005;29:278–80. imaging.computed tomography JComput perfusion stenosis. Neurosurgery. 2004;55:63–7;discussion 7–8. inpatients with carotidacetazolamide reactivity artery test. Neuroradiology. 1991;33:217–22. andacetazolamide studiedwithstableXE-CT capacity resting cerebral blood flow and cerebrovascular reserve Shiroyama Y, etal. on The effect ofEC-IC bypass surgery sion 60–3. cerebral ischemia.Surg Neurol. 2003;59:455–60;discus cognitive inpatientswithhemodynamic impairment Cikrit DF,Cikrit SG,Sawchuk RW, AP, DalsingMC,Lalka Burt Yamashita T, S, S,Nakano Takasago Kashiwagi T, S, Abiko Powers WJ, Clarke WR, GrubbRLJ, Videen TO, Adams HPJ, ------EANM Chapter 7 PET/CT in Radiotherapy Planning of Brain Tumours Roberto Delgado-Bolton, Adriana K. Calapaquí-Terán and Javier Arbizu

Introduction best available in terms of diagnostic efficacy Brain tumours can be classified, according to (sensitivity, specificity, positive and negative their origin, into primary (those originating predictive values) and must have high spa- from the central nervous system) or secondary tial resolution. CT and MRI comply with these (those originating from other tissues and requirements [2]. However, a high contrast metastasising to the central nervous system). resolution (i.e. the ability to differentiate Primary brain tumours include low-grade pathological from normal tissues) is also and high-grade malignant lesions according essential. This is where PET enters the picture. to the World Health Organisation, the most frequent being meningioma (30%) and Why add PET/CT to radiotherapy planning glioblastoma multiforme (20%). of brain tumours? PET/CT in radiotherapy planning For patients with high-grade brain tumours, CT and MRI have a high spatial resolution the ideal treatment is surgery (complete re- and display the anatomy and morphologi- section of the tumour, if possible preserving cal changes in great detail. Their main draw- neurological function) followed by adjuvant back is their low contrast resolution, as a chemotherapy and radiotherapy. The utility consequence of which it is not possible to of radiation therapy has been demonstrated. differentiate between viable tumour tissue Technological advances are increasingly im- and surgery-, radiotherapy- or chemotherapy proving the efficacy of the treatment (mea- -related changes (pseudo-progression, sured by survival) based on increased treat- pseudo-regression) [3]. In contrast, PET has ment precision (reducing the radiation dose a high contrast resolution, and by supply- to healthy tissue and the associated toxicity) ing functional information on the biological without any reduction in the radiation dose and molecular characteristics of the tumour to the treatment volume. it helps in differentiating viable tumour from treatment-related changes, thereby improving Radiotherapy planning requires multidisci- treatment planning for high-precision radio- plinary collaboration and a sequential pro- therapy. Many studies have analysed the role cess that begins with a tumour board deciding of different PET tracers for gross tumour vol- to irradiate a tumour and ends with the ap- ume (GTV) delineation. plication of a treatment plan signed by a board-certified radiation oncologist [1, 2]. A PET tracers key aspect is the appropriate selection and 18F-Fluorodeoxyglucose (18F-FDG) is the delineation of the target volumes and organs main PET tracer used in oncology as it pro- at risk, which is typically done by a radiation vides useful information in many tumours. oncologist on contrast-enhanced CT and/ However, it has been shown to be of limited or MRI. The images used need to be the value in the radiotherapy planning of brain

92 Chapter 7 PET/CT in Radiotherapy Planning of Brain Tumours

tumours [3, 4]. The main issue with 18F-FDG PET/CT technical requisites is the low contrast between viable tumour In order to maximise the potential benefits and normal brain tissue. 18F-FDG uptake is of incorporating the metabolic information increased in anaplastic areas [3]. provided by PET, it is crucial to be aware of all the factors that must be taken into ac- Radiolabelled amino acids (AA) such as 11C- count to guarantee that the images acquired methionine (MET), 18F-fluoroethyl-tyrosine comply with the requirements for treatment (FET) and 18F-fluoro-L-dopa (FDOPA) are the planning and treatment delivery [6, 7]. These main PET tracers used for the characterisa- factors are: tion, therapy planning and follow-up of brain tumours, and all three of the aforementioned 1. Initial patient positioning. Accurate posi- tracers share similar characteristics regarding tioning that ensures consistency and daily uptake intensity and distribution in brain tu- reproducibility of treatment is essential when EANM mours. Normal cerebral tissue has a relatively delivering high doses to a tumour, as an im- low uptake of AA PET tracers as compared portant limitation is the tolerance level of the to tumour cells, which incorporate them at surrounding normal tissue [7]. To ensure the a high rate; furthermore, uptake of AA PET comfort of the patient during all treatment tracers is not dependent on disturbances of sessions, the initial evaluation must include the blood-brain barrier. the physical status and limitations as well as the psychological impact (claustrophobia). In Tumour target volumes defined using AA order to achieve accurate alignment, an im- PET differ from those outlined with MRI using mobilisation device is used in conjunction standard sequences such as T1-weighted with reference ink or tattoo marks [7]. Posi- after gadolinium contrast enhancement (T1- tioning includes: (a) removing all clothing Gd), both in size and in geometrical extension from the upper body to promote comfort [5]. In fact, the tumour target volume as and ensure daily reproducibility; (b) removing defined on T1-Gd MRI seems insufficient to dentures or implants and instructing the include the real biological tumour extension patient not to chew anything during the of high-grade brain tumours. treatment; (c) supporting the head to prevent or limit involuntary movements.

93 1A 2A

2B

1B

Figure 2 A, B: Positioning of a patient with the laser light system

Figure 1 A, B: Table top with a neck support 3. Laser lights. Lateral and sagittal lasers must be used to ensure accurate alignment and 2. Table top. The table top must be flat, nar- positioning. The laser light system installed row and rigid and should allow registration in the PET/CT unit must be in accordance or indexing of immobilisation devices [7]. with the one installed in the radiotherapy Figure 1 presents a table top with a neck unit. Quality controls of the laser lights of support. the PET/CT system must be done routinely to maintain consistency with the treatment unit [7]. Figure 2 shows positioning of a pa- tient with the laser light system.

94 Chapter 7 PET/CT in Radiotherapy Planning of Brain Tumours

Figure 3A–D: Immobilisation systems that must be individualised include thermoplastic masks. The mask is initially rigid (A) but can become pliable when placed in warm water (B). This permits the mask to be moulded to the head and shoulders of the patient (C). It can be attached to the treatment table once the material has cooled (D) 3A 4. Immobilisation. The use of immobilisation devices is essential to deliver high doses of radiation to the treatment volume while

ensuring the protection of critical organs EANM (e.g. crystalline lens, optic chiasm, pituitary gland). Immobilisation systems must be in- dividualised for each patient and should be anchored to fastening systems which in turn 3B must be fixed to the treatment table. For brain tumours, immobilisation systems in- clude: (a) neck supports (standard, frames or vacuum bags) (Fig. 1) and (b) thermoplastic masks which are initially rigid (Fig. 3 A) but become pliable when placed in warm wa- ter (Fig. 3 B), allowing moulding of the mask to the head and shoulders of the patient (Fig. 3 C) and attachment to the treatment table once the material has cooled (Fig. 3 D).

3C

3D

95 When does radiotherapy planning with for radiotherapy planning, and an additional PET/CT have an added value? CT will not be performed. AA PET/CT is of specific value for delineation of the GTV in high-grade gliomas, especially Technical approach to patient positioning those that are non-contrast enhancing for PET/CT for radiotherapy planning on T1-Gd MRI, and low-grade tumours Multidisciplinary collaboration between the that are ill defined on MRI. Additionally, nuclear medicine and radiation oncology as already mentioned, it is of great value units is essential. The patient must be posi- in differentiating between treatment- tioned using the technical instruments men- related changes (pseudo-progression and tioned above in the section “PET/CT techni- pseudo-remission or pseudo-response) and cal requisites”. residual/recurrent tumour [3, 8]. PET/CT with radiolabelled AA or DOTATOC can be used Quality control of image fusion of PET/CT for GTV delineation in meningiomas and and planning CT for tumour delineation glomus tumours. Further research is needed Quality control of image fusion must be done regarding the role of PET with other tracers, routinely, ensuring the patient has been po- such as 18F-fluorothymidine (FLT) to evaluate sitioned adequately in accordance with the proliferation, 18F-fluoromisonidazole (FMISO) protocols. to assess hypoxia and 18F-galacto-arginine- glycine-aspartic acid (RGD) to analyse PACS/RIS and software issues angiogenesis and peptide expression [3,8]. DICOM must be transferred from the PET/CT to the radiotherapy unit. How should it be done? Procedure, technical aspects and quality control Who should participate? Multidisciplinary PET/CT procedure collaboration 18F-FDG PET/CT must be performed in The nuclear medicine and radiation oncology accordance with the EANM guidelines [9, 10]. units must collaborate closely when planning When radiolabelled AA analogues are used the radiotherapy and delineating the instead of 18F-FDG, the guidelines dedicated volumes to treat. The relevant protocols must to these must be followed [11]. If the be strictly followed to achieve the maximum PET/CT data are used for radiation planning, accuracy, consistency and reproducibility. the examination should be performed in the The PET/CT study should be reported in position which will be used for radiotherapy, accordance with established guidelines employing the same dedicated positioning [9−11]. Volume delineation is done taking devices as are used in the radiotherapy into account both MRI and PET. The GTV department [9]. The CT part of PET/CT must based on PET is added to the GTV based on have special characteristics as it will be used MRI and they are fused with the CT (Fig. 4).

96 Chapter 7 PET/CT in Radiotherapy Planning of Brain Tumours

A C

B D EANM

Figure 4 A–D: Radiotherapy planning of a high-grade glioma using MRI and 11C-MET PET/CT: T1-weighted MRI (A), non-contrast-enhanced CT (B), T2-weighted MRI (C) and 11C-MET PET (D). T2-weighted MRI shows an extensive area of enhancement (volumes 1 and 2) that could be due to active tumour and/or oedema. T1-weighted MRI shows a small area of enhancement (volume 2) and an area that does not enhance (volume 1). CT shows an area of changes (volumes 1 and 2). If only MRI or CT were available, the treatment volume would have included both areas (volumes 1 and 2). However, 11C-MET PET evidenced high uptake (tumour) in volume 1 but no uptake at all (non-tumour) in volume 2. Because of this information, the treatment volume included only volume 1, sparing unnecessary toxicity in volume 2

97 What must be taken into account? Conclusion Radiation protection issues: focus on Accurate position reproducibility to ensure radiotherapy workers daily reproducibility of treatment is essen- When radiotherapy planning is done on tial when delivering high doses to a tumour. PET/CT, radiation protection for radiotherapy Multidisciplinary collaboration also plays a workers must take into account their close key role in patient positioning. contact with patients who have been injected with PET tracers.

Patient scheduling Patients referred for radiation treatment planning must be scheduled at the end of the morning or afternoon, after the standard PET/CT studies, in order to cause the least possible delay to the daily routine. These studies entail further complexity and time and therefore must be scheduled last.

98 4. 1. References Chapter 7 6. 5. 2. 3. References Oncol. 2004;73:261–3. unravelingwithout theuseofPET: themyth. Radiother Gregoire V. Is there any future in radiotherapy planning Imaging. 2012;39:771–81. Mol Eur JNuclMed ofintegrationassessment: patterns intherapy planning. of gliomaswithsequentialMRIand¹¹C-methioninePET E, J,Arbizu Tejada JM, S,Marti-Climent 1998;41:989–95. OncolBiolPhys.ning ofmalignant JRadiat gliomas. Int for the3-Dradiationtreatmentoxyglucose PET plan P ticularly exquisiteticularly test or pending and experimental Gregoire V, inradiotherapy planning:Par ChitiA.PET 2010: vol 10, issue1. 83.JICRU. Oxford:ICRU Report Oxford Press. University In: intensity-modulated photon-beam therapy (IMRT). Measurements. Prescribing, recording andreporting ICRU. Unitsand Commission onRadiation International Gross MW, Weber WA, Feldmann HJ, Oncol.ning ofbraintumours. 2010;96:325–7. Radiother Grosu A-L, Weber WW. for radiationtreatment PET plan Oncol.tool? 2010;96:275–6. Radiother , Schwaiger M, Quincoces G,etal. Quantitative volumetric analysis Molls M. Molls The valueofF-18-fluorode Diez-Valle R, Bartenstein Bartenstein Prieto - - - - 99 10. 7. 11. 9. 8. Varrone A,Asenbaum S, Vander Borght T, Imaging. 2015;42:328–54. Mol Nucl Med dure guidelinesfor tumourimaging: version 2.0.Eur J brain imaging using[ F Mol Imaging. 2009;36:2103–10. Mol aging. 2006;33:1374–80. - labelled aminoacidanalogues. Im Mol Eur JNuclMed procedure guidelinesfor braintumourimaging using Vander Borght T, etal. EANM K, Asenbaum S,Bartenstein F R,Oyen Boellaard R,Delgado-Bolton WJG, Oncol.tics. 2010;96:280–7. Radiother ment planning – Synthesis- and biological characteris inradiationtreat radiopharmaceuticals Haubner R.PET 2010;96:298–301. quisition in radiotherapy planning. Oncol. Radiother Coffey M, ac A. Vaandering Patient setup for PET/CT , , Tatsch K Någren K , Eschner W , et al. EANMprocedure guidelinesfor PET 18 , et al. FDG PET/CT: EANM proce et al. FDG PET/CT: F]FDG, version 2.Eur JNuclMed Booij J Booij Giammarile Giammarile , Nobili - - - EANM Chapter 8 PET/MRI for Brain Imaging Peter Werner *, Torsten Boehm *, Solveig Tiepolt, Henryk Barthel, Karl T. Hoffmann, Osama Sabri

Current clinical applications of PET/MRI in Further, it is used to detect specific atrophy neurology patterns, white and grey matter degenera- Magnetic resonance imaging (MRI) is pivotal tion and the implications of some metabolic in the work-up of the majority of neurologi- changes [1]. However, structural changes cal and psychiatric diseases since it provides that are specific to neurodegenerative dis- morphological and functional information eases are usually preceded by months to about the brain. Positron emission tomog- years of characteristic molecular and func- raphy (PET), on the other hand, is able to tional alterations that can be detected with detect and quantify a variety of pathophysio- multi-tracer PET exclusively [2]: logical processes with very high sensitivity. In patients with symptoms of cognitive impair- 1. Characteristic impairments of brain ment or movement disorders, for example, glucose metabolism can be detected by high-resolution and multi-contrast brain MRI [18F] fluorodeoxyglucose (FDG) PET and is essential to rule out non-neurodegenera- allow the early diagnosis of dementia dis- tive causes such as tumours, haemorrhages, orders while MR findings are still absent or ischaemic lesions or inflammatory diseases. unspecific (Fig. 1A).

A

T2 FLAIR Fusion [18F]FDG PET (Glucose metabolism)

Figure 1 A–C: Multi-tracer brain PET/MRI in neurodegenerative diseases. (A) A 61-year-old man with progressive aphasia and cognitive impairment; T2-weighted fluid attenuation recovery (FLAIR) MRI (left) shows enlarged ventricles, mainly left cortical atrophy and white matter lesions. Simultaneous [18F]FDG PET revealed severe left cortical hypometabolism (arrows, middle), indicating a form of frontotemporal dementia as underlying disease.

2. Amyloid PET can accurately detect brain in which the presence of amyloid deposits amyloid load in Alzheimer’s dementia (AD), ultimately leads to cognitive decline (Fig. 1B).

* Contributed equally 100 Chapter 8 PET/MRI for Brain Imaging

B

[18F]Florbetaben PET T2 FLAIR Fusion (β-Amyloid)

Figure 1 A–C: Multi-tracer brain PET/MRI in neurodegenerative diseases. EANM (B) A 44-year-old woman with cognitive impairment and a family history of Alzheimer’s dementia (father). Brain MRI (left) was unremarkable but the presence of intense β-amyloid depositions in the whole brain grey matter on [18F]florbetaben PET (arrows) confirmed the clinical suspicion of AD.

3. [18F]DOPA PET can specifically visualise in specialised centres to diagnose parkinso- dopaminergic neurotransmission in the nian syndromes [3] (Fig. 1C). basal ganglia and is used in clinical routine

C

T2 TSE Fusion [18F]DOPA PET (Dopamine metabolism)

Figure 1 A–C: Multi-tracer brain PET/MRI in neurodegenerative diseases. (C) A 76-year-old woman with unspecific movement disturbance of the legs. T2-weighted turbo spin echo (TSE) MRI was unremarkable; moreover [18F]DOPA PET revealed age-appropriate dopamine synthesis in the basal ganglia (arrows). Thus, PET/MRI allowed the exclusion of neurodegenerative Parkinson’s disease as the underlying disorder

101 Recently, an integrated amyloid PET/MRI ex- to decide whether surgery or biopsy of cra- amination algorithm has been proposed to nial masses is necessary (Fig. 2A). In addi- deliver a maximum of imaging parameters, tion, amino acid PET delivers reliable tumour necessary for the diagnosis of a wide range volumes for radiation or surgery planning in of dementia disorders [1]. Conventional MRI situations in which the tumour itself cannot is also essential in the clinical management be distinguished from unspecific surround- of brain tumours. With a variety of contrasts ing MRI signal abnormalities (e.g. due to oe- it provides information about the type of dema) (Fig. 2B). Lastly, somatostatin receptor tumour and its size, extension and grading. PET can highly selectively visualise tumour Moreover, (a) enhancement of areas after ap- tissue arising from the meninges (Fig. ). Si- plication of a contrast agent (e.g. Gadovist) multaneous PET/MRI also plays a role in basic allows an estimation of blood-brain barrier research into brain function and pathological breakdown, which is typical of high-grade mechanisms in neurological diseases such as malignant brain tumours, (b) dynamic sus- stroke, dementia and addiction. The combi- ceptibility contrast perfusion measurements nation of functional MRI, contrast-enhanced improve differentiation between recurrent MRI and multi-tracer PET allows the simul- tumour and radiation-induced necrosis and taneous assessment of rapidly fluctuating (c) MR spectroscopy provides information brain signals like brain perfusion, transmitter about malignant proliferation and promising release or the occupancy of specific recep- targets for eventual tumour biopsy. tors in the brain. Related examination pro- tocols exhibit a high degree of complexity In various clinical situations, however, MRI is and are usually restricted to special research insufficiently specific and does not deliver facilities (for more information, see Werner et the desired information. Here, simultaneous al. [2]). amino acid PET is helpful, since it reflects the vastly increased amino acid transport into brain tumour tissue. PET is able to dis- tinguish between unspecific changes after tumour (radiation) therapy and tumour re- lapse. Amino acid PET is also helpful in the primary evaluation of unspecific MR findings

102 Chapter 8 PET/MRI for Brain Imaging

A

T2 FLAIR T1 + CE Fusion [11C]Methionine PET (Aminoacid transport)

Figure 2 A–C: Multi-tracer brain PET/MRI in neuro-oncology. (A) A 60-year-old woman with unspecific T2 fluid attenuation recovery (FLAIR) signal alteration, without contrast enhancement (CE) on the T1 contrast image (arrow). Moreover, amino acid transport EANM assessed by [11C]methionine PET was unremarkable (right). A highly malignant glioma could be excluded and, due to the lack of symptoms, follow-up imaging was chosen over biopsy.

B

T2 FLAIR T1 + CE Fusion [11C]Methionine PET (Aminoacid transport)

Figure 2 A–C: Multi-tracer brain PET/MRI in neuro-oncology. (B) In this glioma patient the extension of signal alterations on the T2 FLAIR and T1 CE images (red lines) differed substantially. [11C]Methionine PET identified highly active tumour areas with increased amino acid transport (right).

103 C

T2 FLAIR T1 + CE Fusion [68Ga]DOTATOC PET (Somatostatin Receptors)

Figure 2 A–C: Multi-tracer brain PET/MRI in neuro-oncology. (C) Somatostatin receptor imaging with the highly selective PET radioligand [68Ga]DOTATOC traced tumour tissue in this 74-year-old woman with a meningeal brain tumour (arrows). On MRI the contrast between tumour and healthy tissue was lower

Advantages and limitations to be mentioned, however, that in order to Even before the introduction of hybrid PET/ obtain quantitative PET data the correction MRI, the combination of sequentially ac- of the photon attenuation by head tissue is quired brain MRI and PET information of- necessary and that, unlike in PET and PET/CT, ten proved superior to the single-modality this attenuation information is not gathered approach in the diagnosis and work-up of by CT or dedicated transmission scans but neurological diseases. Although favourable, indirectly by MRI. An associated drawback is this imaging approach has not been ap- the fact that the attenuation correction (AC) plied consistently in clinical routine, mainly algorithms currently implemented in PET/ due to organisational issues and reimburse- MRI systems do not consider cortical bone, ment questions. It may be hypothesised which leads to significant underestimation of that hybrid brain PET/MRI, simply by ensur- the AC-PET signal. New AC approaches ad- ing consistent availability of both modali- dressing this important issue have been de- ties and their joint interpretation by trained veloped but have not yet been implemented physicians, may improve diagnostics in this in the commercially available systems (May field [2]. Compared with sequential MRI and 2015). Moreover, running a hybrid PET/MRI PET(/CT) a hybrid examination has other ob- requires technologists who are specifically vious advantages, such as improved patient trained in both modalities. Nuclear medi- comfort (through the scheduling of fewer cine technologists are trained in handling examinations), reduced overall examination ionising radiation and are acquainted with time and reduced radiation exposure. It has time-critical scheduling of examinations

104 Chapter 8 PET/MRI for Brain Imaging

with short-lived radiotracers. In contrast, hearing protection, earplugs or earphones MRI technologists are trained in MR safety should be chosen over headphones since and skilfully operate the MR scanner to pro- they cause less attenuation of the PET signal. vide high-quality diagnostic images and mi- The upper part of the head and neck coil is nimise artefacts in a large variety of different closed and the patient is instructed regard- sequences and protocols. For combined PET/ ing the emergency bell. It is possible to at- MRI, a basic understanding of the underlying tach a mirror to the coil which shows the physics of both modalities is required in the window to the control room in order to avoid first instance in order to be able to adapt ac- the otherwise oppressive feeling when look- quisition parameters during a patient scan. ing at the ceiling of the gantry. Due to the To run a hybrid PET/MRI with a small staff, relatively low room temperature at a PET/MRI dual-trained technologists are desirable but facility, use of a blanket to cover the patient such technologists are extremely rare. A joint is recommended. Finally, the head centring EANM statement of the Society of Nuclear Medicine is done via laser positioning, which should and Molecular Imaging Technologist Section only be activated once the patient’s eyes are and the Section for Magnetic Resonance closed. With a strongly reclined head the pa- Technologists has underlined the need for tient is moved into the gantry until the laser advanced-level educational programmes for cross is a little below the nose and the cen- PET/MRI technologists [4]. tre position is then confirmed by pressing a button. Thereafter, the single MR sequences Technical background and workflow need to be defined and the PET scan is aspects planned; care has to be taken that the entire Positioning and preparation brain is within the field of view. While the PET All patients receive intravenous access and acquisition times differ depending on the are repeatedly asked about contraindications specific tracer, the following reconstruction (e.g. ferromagnetic metal implants, pace- parameters have generally been found to be makers). Due to the strong magnetic field, optimal for brain imaging on a Siemens mMR patients need to remove all metal objects PET/MRI system: 8 iterations, 21 subsets, Y and dentures before entering the examina- offset = -41, zoom = 2.8, 3.0-mm Gaussian tion room. The patient is then placed on the filter (University Hospital Leipzig, Germany). patient bed as comfortably as possible. This is achieved by use of positioning aids such Once set-up has been completed, the PET/ as knee rolls and foam wedges to stabilise MRI workflow itself is straightforward. The the head. The patient’s shoulder should be start of the PET acquisition is accompanied in contact with the head and neck coil. If the by automatic acquisition of the MR se- head is too big for the coil, the support pad quences for AC. During the PET acquisition, can be removed to create some space. For MRI can be handled flexibly; all subsequent

105 MR sequences are adapted and planned by a non-magnetic syringe shield, means of a three-axis localiser prior to the the space in the gantry is limited and intra- actual MR scan. Additional individual MR se- venous access via a vein on the back of the quences can be planned and acquired as de- hand is helpful in facilitating the manda- sired during the remaining PET scan time or tory tracer application within the system. beyond. Usually, the overall duration of a si- Two MR localisers are acquired first, then a multaneous PET/MRI study is determined by workflow break needs to be implemented in the MR examination time. Due to the distri- the protocol immediately before the actual bution dynamics and the relatively long half- diagnostic MRI and PET scan begins. Upon lives of the radiotracers, they are generally tracer injection within the scanner room by injected prior to the PET/MRI session. The in- the technologist, the actual scan is started by jection of MR contrast medium, on the other a second person sitting at the console; it is hand, is tightly coupled to individual MR se- necessary to write down the application time quences within the hybrid imaging protocol for post hoc calibration of the injected activ- and is performed while the patient is lying in ity. If the tracer is applied during the actual the scanner; however, even with simultane- scan, the tracer activity can only be entered 15 ous injection (e.g. [ O]H2O and gadobutrol into the system afterwards, which makes in research applications in the scanner), no a retrospective reconstruction mandatory. interactions have been reported [2]. At the end of the 10-min PET scan, during which MR sequences can be acquired, the Dual time-point amyloid PET/MRI patient is allowed a break and leaves the By collecting early dynamic (e.g. 1–8 min examination room so that another patient p.i.) data in an amyloid PET scan, informa- can be examined. After about 60–70 min the tion on perfusion as a surrogate of glucose second part of the examination starts, again metabolism and on brain amyloid load can with two localisers. Now the application type be derived with only one radiotracer. When and time can be entered and during the sub- this approach is combined with simultane- sequent 20-min PET scan, acquired from 90 ous MRI, the most important brain imaging to 110 min, the remaining MR sequences can information for the assessment of neurode- be acquired (see Fig. 3 for details). generative disorders can be gathered dur- ing one imaging session [1]. Despite use of

106 Chapter 8 PET/MRI for Brain Imaging EANM

Figure 3: [18F]Florbetaben (amyloid) PET/MRI workflow. Upon injection of the radiotracer, the early 10-min PET scan is started, which is accompanied by the acquisition of the DIXON sequence for attenuation correction (AC) and two more diagnostic MR sequences. After a break of ~70 min, the patient re-enters the scanner and during the late 20-min PET scan more diagnostic MR sequences can be acquired. Note: the DIXON VIBE sequence needs to be acquired again for AC of the second PET scan. Sequence: image plane/slice thickness (mm)/no. of slices/TR and TE (ms)/FOV (mm)/ matrix; 3D Dixon Vibe (DIXON): coronal/3.12/128/3.6 and 1.23/500×300/256; T2 fluid attenuation inversion recovery (FLAIR): axial/3/46/9000 and 93/230/320; T2 turbo spin echo (TSE): axial and coronal/3/50/6000 and 100/230/384; T1 FLASH: axial/3/50/395 and 2.97/230/320; susceptibility weighted imaging (SWI): axial/1/112/25 and 20/230/256; T1 magnetisation prepared rapid gradient echo (MPRAGE): sagittal/1/176/1900 and 2.53/250/256; echoplaner imaging (EPI): sagittal/1/176/1900 and 2.53/250/256. Note: PET reconstruction parameters are provided in the main text

Possible artefacts considered, which leads to a ~10–15% un- As stated above, the PET signal needs to be derestimation of the PET signal in the vicin- corrected for the attenuation of photons ity of cortical bone [5]. Since the attenuation on their passage through the tissue of the values in PET/MRI cannot be measured but head. Current PET/MRI AC sequences clas- rather must be predicted from the MR signal sify three tissue classes: fat, water and air. As by advanced computer algorithms, prob- already noted, bone tissue is currently not lems may arise if there are any deviations

107 with irregular facial bone or skull lesions (e.g. but even without any visible cause the un- due to prior neurosurgery). Hence, a ventricle derlying algorithms sometimes fail to assign shunt system may lead to misclassification of the orrect attenuation values to the head tis- the surrounding brain tissue classes (Fig. 4A), sue classes (Fig. 4B).

Figure 4 A, B: Attenuation correction (AC)-related artefacts. (A) In this patient the MR signal is heavily distorted in the vicinity of the ventricular brain shunt [arrow in T1 with gadolinium (Gad)]. As a consequence, the system fails to derive a correct μ-map from the Dixon VIBE sequence for AC (μ-map) and this area is misclassified as air but not as part of the head. Hence, there is no corrected PET information in the affected voxels available [asterisk in PET(AC)] and the non- corrected PET data need to be evaluated instead [PET(NAC)]. (B) The system wrongly assigned the attenuation values of fat and water (failed μ-map) whereas the correct assignment can be seen below (correct μ-map). Incorrect assignment of photon attenuation values results in an overall overestimation of the PET signal after reconstruction [failed PET(AC)]. The PET reconstructed from the correct Dixon-based μ-map delivers a substantially lower PET signal [correct PET(AC)]

108 3. 1. References References Chapter 8 2. 5. 4.

Barthel H,Schroeter ML,Hoffmann Barthel K-T, Sabri O. PET/ settings. Eur J Nucl Med Mol Imaging 2015;42:512–26. settings. Imaging Mol Eur JNuclMed and future inclinicalandresearch role ofbrainPET/MRI Werner P, O. A, Sabri H, Drzezga Current Barthel status erative diseases. 2014;55(Suppl2):47S–55S. JNuclMed imaging in neurodegenclinical applications of PET/MR 2014;84:206–16. a strong spatialbiaswhenignoring bone. NeuroImage implies MR-basedattenuation correction neurology: AE, Højgaard imaging L,etal. in Combined PET/MR Andersen FL,Ladefoged CN,Beyer T, SH,Hansen Keller Technol 2013;41:108–13. tion for MagneticResonance Technologists. JNuclMed Imaging Molecular Technologist- and the Sec Section and ofNuclearMedicine a jointpaperby theSociety imagingGN, Hunt consensus JKE, et paper: al. PET/MR Gilmore CD, Comeau CR, Alessi AM, BlaineM,El Fakhri 2015;45:224–33. NuclMed Semin MR indementiaandotherneurodegenerative diseases. Drzezga A, Barthel H, Minoshima S, Sabri O. S,Sabri H,Minoshima Potential A,Barthel Drzezga - 109

EANM Chapter 9 Brain Death Marko Grmek

Introduction the availability of the test. Ancillary tests can Brain death (BD) may be defined as the com- shorten the procedure for determining BD plete and irreversible loss of all brain and in some cases because these tests can im- brain stem functions [1, 2]. A person with mediately follow the clinical examination; BD is considered legally dead [3]. In certain this is especially relevant in persons who are cases, for example when treatment with potential organ donors. An ancillary test can mechanical ventilation is performed, some also be used in persons who are suspected of somatic functions such as circulation can be being brain dead but cannot be completely active despite BD. clinically examined, for example those with severe head injuries [7]. When the results of In most European countries, when and how both the clinical examination and the ancil- a person can be pronounced brain dead is lary test match the BD diagnostic criteria, precisely defined [4]. One would expect the an apnoea test can be performed. Ancillary procedures for determination of BD to be studies for determination of BD can also be similar across Europe; in fact, however, they performed in children [8]. differ considerably. In approximately 50% of European countries, the diagnostic proce- Ancillary test results alone without a clinical dure requires two or three clinical tests, with examination do not suffice for a diagnosis of a specified minimum time interval between BD. them, usually 6–24 h [4]. In other countries, an ancillary or confirmatory test is also re- Brain death scintigraphy quired. The apnoea test is usually the final Two types of scintigraphic examination can step in the determination of BD. Within the be used in the procedure for determining United States, the procedures for determin- BD [2, 9, 10]. In the past, scintigraphy of the ing BD also differ considerably [5]. head (the brain) with hydrophilic agents was employed. Such radiopharmaceuticals Ancillary tests [99mTc-pertechnetate, 99mTc-glucoheptonate Legislation and/or the guidelines usually de- and 99mTc-diethylene triamine penta-acetic fine precisely when a certain ancillary test acid (DTPA)] do not cross the blood–brain may be used in the diagnostic process for barrier (BBB). When using hydrophilic radio- determining BD. The following are the most pharmaceuticals, only blood flow images frequently used ancillary tests [6, 7]: cere- of the head should be acquired. Since such bral angiography, electroencephalography images are sometimes difficult to interpret, (EEG), transcranial Doppler ultrasonography this type of examination is now used only and cerebral scintigraphy. Selection of the rarely when determining BD. Instead, the test depends on the state or condition that main diagnostic procedure for this purpose has resulted or could result in BD and on is brain perfusion scintigraphy with lipophilic

110 Chapter 9 Brain Death

radiopharmaceuticals that do cross the BBB, 2. It must be ensured that the radiopharma- e.g. 99mTc-hexamethylpropylene amine oxime ceutical, 99mTc-HMPAO or 99mTc-ECD, is of the (99mTc-HMPAO) and 99mTc-bicisate (99mTc-ECD). appropriate quality. Radiochemical purity should be determined on each vial prior to Brain perfusion scintigraphy injection using the methods outlined in the Brain perfusion scintigraphy enables both package inserts. Radiochemical purity should presentation of the blood flow and perfu- be >90% for ECD and >80% for HMPAO. The sion of the brain [2], making interpretation recommended radiopharmaceutical activ- of examination results considerably easier. In ity is 370–1110 MBq. It is necessary to make line with the SNM Practice Guideline for Brain sure that the correct radiopharmaceutical is Death Scintigraphy, brain perfusion scintigra- applied. phy is primarily used as an ancillary test in de- 3. The radiopharmaceutical is administered termining BD when the presence of certain EANM intravenously with the bolus technique. Such clinical conditions is likely to render clinical a technique allows acquisition of good blood assessment less reliable [11]. Such conditions flow images. include severe hypothermia, coma caused by barbiturates, electrolyte or acid–base 4. The acquisition takes place in two stages. imbalance, endocrine disturbances, drug First, a blood flow study of the head and neck intoxication, poisoning and neuromuscular is performed. Data acquisition begins simul- blockade. BD scintigraphy may also be help- taneously with the administration of the ful in persons who are being considered as radiopharmaceutical. Usually, a one-minute possible organ donors. series of 1- or 2-s images is acquired. Perfu- sion images are obtained during the second The performance of brain perfusion scintig- stage, acquisition commencing 20 min after raphy to diagnose BD differs somewhat from administration of the radiopharmaceutical. its use in patients with neurological condi- SPECT acquisition is most frequently used; tions [12]: only occasionally is planar scintigraphy in multiple projections performed. In both cas- 1. The examination must not further com- es, the entire brain and the brain stem must promise the medical condition of the pa- be visualised. tient. For this reason, brain perfusion scintig- 5. Interpretation of the results of blood flow raphy for determination of BD is usually not studies requires an experienced specialist performed in persons with circulatory insta- who can recognise the presence or lack of bility or when transport or movement could activity in the brain. Interpretation of perfu- further compromise the person’s medical sion images is less complicated. If activity in condition. the brain and the brain stem does not appear

111 on the perfusion images, the examination re- unit to prepare a radiopharmaceutical. The sult is compatible with BD. patient’s ward is asked to prepare the patient for transport to the Department for Nuclear Brain perfusion scintigraphy for Medicine within one hour. The schedule of determination of brain death – our work on the gamma camera is rearranged if experience necessary. New legislation regarding the use of brain perfusion scintigraphy as an ancillary test in Preparation of the radiopharmaceutical: The the process of determining BD was adopted radiopharmacist needs about one hour to in Slovenia in 2001. Prior to that, BD scintig- synthesise stabilised 99mTc-HMPAO and per- raphy was used only rarely for this purpose. form quality control. A radiochemical purity Since 2001, brain perfusion scintigraphy has of at least 80% is required. been performed on 30–40 persons a year in whom the findings of clinical examination Transportation and positioning of the patient were compatible with BD. Between 2001 to the gamma camera: When the radiophar- and 2008, a total of 259 examinations in 253 maceutical has been prepared, the physician persons (174 men and 79 women) were per- asks the patient’s ward to bring over the pa- formed [13]. Patients were on average 41±19 tient. The transportation usually takes around years old; 38 (15%) were younger than 20. 10 min. The patient is carefully moved from Complete discontinuation of brain perfusion the bed to the gamma camera table and his was established in 233 (90%) persons; in the or her head is fastened. remaining 26 (10%), BD was not confirmed because activity was present in a small or Radiopharmaceutical administration: Most large part of the brain. Brain perfusion scin- patients already have an intravenous line tigraphy was repeated after 1–8 days in six inserted. We check whether the intravenous of these patients. Upon re-examination, no line is unobstructed, apply the stabilised activity was visible in the brain in any of the 99mTc-HMPAO, activity 600 MBq (proportion- patients. ately less in children), and flush the system with saline solution. In the Ljubljana University Medical Centre, Department for Nuclear Medicine, brain per- Image acquisition: A blood flow study is fusion scintigraphy in the diagnostic process usually not performed. SPECT acquisition is for determining BD is performed as follows: initiated 15 min after the radiopharmaceutical administration and must encompass the A request for BD scintigraphy: The nuclear entire brain, including the brain stem. The medicine physician who receives a request acquisition is usually performed on a double- for BD scintigraphy asks the radiopharmacy head gamma camera with the following

112 Chapter 9 Brain Death

acquisition parameters: time per view, 20 s; the brain and the brain stem is concordant no. of views, 2×32; matrix size, 128×128; with a diagnosis of BD. zoom, 1.23; orbit, non-circular. The duration of the procedure is approxi- Reconstruction: The quality of the raw images mately 2 h. is examined. If the images meet quality stan- dards, a reconstruction is made. Iterative re- Summary construction is performed, with 12 iterations Brain death scintigraphy is one of the ancil- and 4 subsets. At the same time, the exam- lary tests that can be used in the process of inee is moved from the gamma camera to his determining BD. The most frequently used or her bed. scintigraphic method is brain perfusion scin- tigraphy with 99mTc-HMPAO. The procedure

Interpretation and reporting: Assessment of for and the interpretation of this examination EANM the scintigrams that have been obtained is are relatively uncomplicated. straightforward (Fig. 1). Absence of activity in

A

113 B

Figure 1A, B: Brain perfusion scintigraphy in determining BD. In the first investigation (A), activity is present in the brain and the criterion for BD is not fulfilled. The investigation was repeated on the same patient after 7 days (B). Absence of activity in the brain and brain stem confirms the cessation of brain circulation – the result is compatible with BD

114 7. 6. 3. 2. 1. References References Chapter 9 5. 4.

Guidelines for the determination of death. Report of Guidelines for ofdeath.Report thedetermination versies. NuclMed. 2008;38:262–73. Semin concepts,mination of brain death: criteria, and contro J. studiesinthedeter LS,Kolano Radionuclide Zuckier the definitionofbraindeath.JAMA. 1968;205:337–40. to School examine Medical Committee oftheHarvard ofthe Ad Hoc A definitionofirreversibleReport coma. Wijdicks EF, Varelas PN, Gronseth GS, Greer DM. brain_death_guidelines.htm professionals/hospital_administrator/letters/2011/ death inadults. Neurology. 2010;74:1911–8. brain guidelineupdate: Determining dence-based Neurology.tion: Acallto action. 2013;81:2009–14. termining braindeath.2011.http://www.health.ny.gov/termining State Task Force onLife andtheLaw. Guidelinesfor de F, Ardelt inbraindeathdetermina A.Practice variability Shappell CN,Frank SanchezM,Goldenberg K, JI,Husari for asolution.Neurocrit Care. 2014;21:376–82. inEurope: inbraindeathdetermination ability looking IA,Bronco G,Crippa Citerio A, Vargiolu A,SmithM. research. JAMA. 1981;246:2184–6. Problems inmedicineandbiomedicalbehavioural the President’s Commission for theStudyofEthical the medicalconsultantsondiagnosis ofdeathto New New York of Health and New State Department York Vari Evi ------115 8. 8. 13. 12. 11. 10. 9.

Nakagawa Nakagawa TA, Ashwal M,MysoreMR,BruceD, S,Mathur tion. Eur J Nucl Med Mol Imaging. 2009;36:S389. Mol tion. Eur JNuclMed scintigraphy for- brain perfusion braindeathconfirma M,FettichGrmek J, Groselj M.Useof C,Dolenc-Novak 2009;36:2093–102. maceuticals, Imaging. version Mol 2.Eur JNuclMed using99mTc-labelled SPECT perfusion radiophar T, etal. Någren EANMprocedure K, guidelinefor brain OL,NobiliF,Kapucu Varrone J, A,Booij Vander Borght scintigraphy 2.0.JNuclMed Technol. 2012;40:198–203. G, NagelJS,etal. Donohoe KJ, Agrawal G,Frey KJ, Donohoe Gerbaudo KA, VH, Mariani NuclMed. 2012;42:27–32. death. Semin Sinha P, Conrad ofbrain GR.Scintigraphic confirmation NuclMed. 2003;33:312–23. of braindeath.Semin Conrad GR, SinhaP. test Scintigraphy asaconfirmatory 2011;39:2139–55. the 1987 Task Force recommendations. Care Crit Med. of brain death in infants and children: an update of Conway EEJr, etal. Guidelinesfor thedetermination SNM practice guidelineforSNM practice brain death - EANM Chapter 10 Health Care in Patients with Neurological Disorders Claudiu Peștean

Introduction they start with those actions designed to All medical activities are focussed on pa- assist in early detection of disease and end tients. Even in the contexts of medical re- with patient education during the recovery search, the medical industry and medical process, which is very important in helping care, every effort made by those involved is the patient to maintain his or her recovered ultimately intended to benefit patients. Nurs- health [1]. ing aims to meet patients’ needs but also to promote their state of health, to prevent ill- Nursing care ness, to restore health and to help the patient Nursing care consists in the nursing process. to deal with illness or even death. The nursing process should be constructed on the basis of a very strong relation with From a nursing point of view, health pro- the patient and with the family or caregivers, motion consists in all those actions that in- and patients should be partners in the de- fluence the lifestyle of the patient in order sign and implementation of the process [1]. to enhance quality of life. It entails not only Nursing care involves a holistic approach to measures that help in avoiding illness, but patients and their problems, as is underlined also steps to encourage the patient to be by most papers on nursing theory [2]. It is aware of his or her health status and to teach considered simultaneously to be both an art behaviours that can contribute in maintain- and a science [1]. In order to provide effective ing or improving health. The provision of nursing care, it is necessary to have special pertinent information and appropriate use of skills such as determination, empathy, critical referrals are important elements in this pro- thinking abilities, and comprehensive medi- cess [1]. cal knowledge.

Prevention of illness means to reduce risks At the same time, the nursing process is a and to maintain optimal functioning of the systematic and accurate method for provi- individual. In addition, early detection of ill- sion of suitable medical care. Theoretical ness is essential. For these purposes, educa- knowledge and practical skills are deployed tional programmes are needed, as well as with the aim of developing a strategy that health assessments at different levels and in will be most effective in helping the patient different areas of interest [1]. to recover health and also in teaching the pa- tient how to maintain this state in collabora- By definition, in the ill patient these aims tion with those close to him or her and taking have not been accomplished, so the first pri- into account all the environmental factors ority is to assist in restoration of health. The that influence the patient’s life. The nursing specific activities employed to achieve this process starts as all systematic methods start, end represent the core of nursing practice; with accurate documentation and a research

116 Chapter 10 Health Care in Patients with Neurological Disorders

step, the so-called nursing assessment. Once The typical standardisation of the nursing the assessment has been completed, the process involves four main steps: assess- practitioner will establish the nursing diag- ing, diagnosing, outcome identification and nosis, identifying the actual but also the po- planning, and evaluating (Fig. 1). tential health problems. With these problems as input data, the nursing plan is established Assessing and applied, with inclusion of all nursing During the assessment, data are collected re- interventions to be imposed. Based on this, lating to the patient’s health status, ability to one can conclude that nursing care is an in- manage his or her healthcare and the need dividualised process adjusted according to for nursing care. In this phase a pertinent and each patient’s needs. In the final part of the comprehensive database is created which process – evaluation – the effectiveness of provides the starting point for the nursing the implemented plan is assessed in order to plan (Fig. 1). Assessment is not limited to the EANM determine the achievement of goals. Bear- initial part of the nursing care process; rather ing in mind these steps, the nursing process it should be ongoing because the health sta- can be characterised as a systematic process, tus of the patient may change quickly, new comprising an ordered sequence of activi- information may be obtained and some in- ties. It is also a dynamic process because of formation may no longer be relevant owing the extensive interaction and the overlap be- to changes that occur during the attainment tween the nursing steps. of cure. The person who assesses the health status of patients will collect the data dur- The nursing process is interpersonal in na- ing a comprehensive interview designed to ture, being based, as already mentioned, gather information on the patient, the fam- on the complex relation between the prac- ily, and the environment, thereby providing titioner and the patient and being patient a holistic perspective. The information is then focussed and not task centered [1]. While the prioritised according to the patient’s imme- process is focussed on the patient, it is out- diate needs or how the interviewer antici- come oriented because its aim is to achieve pates the situation developing. Furthermore, the goals identified as benefitting the pa- the assessment is not restricted only to the tient; this is very important because knowing interview. It is also based on data collected the outcomes that may be expected, nursing using an observational method, i.e. the per- actions can be prioritised and nursing care son who assesses the patient will analyse the successfully provided. patient’s condition, and on the results of evi- dence-based techniques such as the nursing The nursing process should be universally examination and measurements of specific applicable and able to respond to a wide medical parameters. variety of individuals/patients and situations.

117 The recorded data are critically evaluated to use a body systems approach to organise to detect all possible forms of bias, possible assessment data in order to better detect abnormal findings, possible inconsistencies, nursing problems [1]. The nursing diagnosis and missing information. The data will be comprises all the identified problems that the structured upon significance-based criteria. nurse can treat independently. These prob- Usually during the initial assessment the fo- lems may change from day to day in accor- cus is on a specific problem. It is also possible dance with the patient’s response to illness to encounter physiological or psychological and the provided health care. Usually the emergencies (as often occurs in neurology) written nursing diagnosis has two elements: and in these circumstances the assessment the patient’s problem and its cause, i.e. the will concentrate on life-threatening prob- etiology; in addition, a third element can be lems. When the nursing process continues included, namely the characteristics of the over a longer period, the initial assessment problem [1]. The nursing diagnosis should be may be followed by further assessment activ- critically evaluated and validated. When ap- ities designed to evaluate the patient’s status propriate, the nursing diagnosis may also be and compare it with the initial one. This is not validated by the patient. This is one more ex- so relevant in diagnostic nuclear medicine ample of how the patient is a partner in the for neurological disorders, but it could be ap- nursing process, not merely a client. plied as a phase in the nursing process im- plemented by the nursing team who assess Outcome identification and planning the patient in the neurology department and After the health problems have been identi- send the patient for specific nuclear medi- fied and prioritised, the nursing actions can cine procedures. The collected data may be be planned (Fig. 1). In this stage the out- subjective, taken from the patient’s perspec- comes will be established and the appro- tive, or objective, i.e. measurable and observ- priate nursing interventions will be planned able. Both types of data should be collected according to nursing theory and evidence- because they complement one another [1]. based nursing guidelines. The nursing plan will be communicated to the patient and to Diagnosing the health care team. The nursing plan aims Upon completion of data collection, all the to prevent, solve or reduce a patient’s health information is analysed and synthesised. A list problem and to achieve the patient’s expec- of suspected problems, potential problems tations relating to health [1]. and risk factors is created and resources and strengths are identified as the basis for health During this step of the nursing process the promotion and health recovery actions [1] first intention is to identify the expected out- (Fig. 1). This second step of the nursing pro- comes, which are individualised on the basis cess is termed nursing diagnosis. It is good of the patient’s characteristics and condition.

118 Chapter 10 Health Care in Patients with Neurological Disorders

When identifying the anticipated outcomes, team to deliver goal-oriented, pertinent the first partner is the patient and then the nursing care and will assist in standardisation family, who can subjectively formulate their of the activities requested to help the patient. expectations. Starting from the diagnosis, outcomes may be derived in order to re- Implementing spond to the patient’s needs. The outcomes In this step of the nursing plan, all the activi- will be identified on the basis of clinical ex- ties planned on the basis of the information pertise, scientific and medical evidence, gathered in the previous steps should be associated risks, costs and benefits, the pa- implemented by the nursing team (Fig. 1). tient’s values and ethical considerations. The time points by which these outcomes should The nurse will have the role of health care be met are then established. The identified coordinator. Patients are seen by various spe- outcomes may be modified if this is necessi- cialists during the provision of medical care EANM tated by changes in the status of the patient. and each of them focusses on different as- All the above will be documented [1]. pects. The nurse needs to summarise all the findings and recommendations from all the In order to meet the expected outcomes, the healthcare professionals and to prepare pa- patient will need professional help. Nursing tients and their families to participate maxi- interventions must be established that are mally in the plan of care. The nurse may also appropriate in the context of the nursing have an important role in liaising between diagnosis. These interventions must take the members of the medical team. When im- account of nursing abilities, available time plementing the nursing plan, critical thinking and resources and must be safe, efficient, re- is necessary: nursing personnel should iden- alistic, in accordance with research findings tify the best means of implementation and and standards of care, compatible with all assess whether the interventions are indeed the patient’s values and beliefs, acceptable supported by research findings in the field of to the patient and compatible with other interest. If, during monitoring, new evidence therapeutic activities. Nursing interventions arises relating to the patient’s response, the may be nurse initiated, when no physician’s plan should be modified accordingly [1]. order is needed and the intervention is based only on the nursing diagnosis, or physician Evaluating initiated, when the nursing personnel imple- Evaluation of the achieved outcomes estab- ment interventions ordered by the physician lished within the nursing plan is very impor- in response to the medical diagnosis. tant since this step assesses the success of nursing care and the need to correct, change, The nursing plan should be documented improve or continue the nursing care activi- in written form, which will help the nursing ties (Fig. 1). It is important that evaluation is

119 performed as early as possible; postponing very poor. Initial signs and symptoms are in- it until just before patient discharge can lead sidious and include memory loss. There may to failure to implement modifications in a also be difficulty in learning and memorising timely manner. The evaluation is based on new information, difficulty in concentrating various criteria and standards designed to and impaired hygiene. Gradually, progression assess the accomplishment of desired out- to severe signs and symptoms occurs, with comes. Final data are collected by nursing impairments of linguistic and spatial abili- personnel in order to evaluate the success ties and appearance of complex symptoms. of the nursing plan. If all the outcomes have These include difficulty in abstract thinking, been achieved and no additional nursing di- difficulty in performing activities that require agnostics are identified, the nursing process judgment, deterioration in motor function may be closed. Very often, however, the pro- and coordination, impaired ability to read cess must be continued owing to failure to and write, disorientation and performance of achieve the established outcomes. The nurs- repetitive actions [3, 4]. ing plan then needs to be reviewed and re- vised as necessary; this revision may include The nursing personnel will focus on support- the deletion or modification of nursing diag- ing the patient to maintain present abilities noses, alteration of the expected outcomes and to compensate for those abilities that so that they are more realistic, adjustment of have been lost. The patient and his or her the time points for achievement of specified family will be assisted in establishing an ef- outcomes and alteration of nursing interven- fective communication system according tions [1]. to the altered cognitive abilities. Emotional support will be offered. The patient will be Neurological care protected from injuries by providing a safe In this section we shall review those neuro- environment. The patient will be encouraged logical disorders more often addressed in and taught to perform physical exercises to diagnostic nuclear medicine departments maintain physical abilities. Appropriate infor- using specific radiotracers described earlier mation about the disease will be provided to in the guide. The main clinical characteristics the patient and to the family [4]. of each disorder and the appropriate nursing interventions will be summarised. Parkinson’s disease The early symptoms of Parkinson’s disease Alzheimer’s disease are non-specific and include slower walk- Alzheimer’s disease is a common neurologi- ing and dressing and non-motor symptoms cal disorder that affects a high percentage of such as disturbed sleep, and depres- elderly persons and accounts for more than sion. The most obvious symptoms are bra- half of cases of dementia. The prognosis is dykinesia (lack of spontaneous movements,

120 Chapter 10 Health Care in Patients with Neurological Disorders

lack of facial expression, reduced arm swing), seizure [3, 4]. Depending on the localisation rigidity and rest tremor; although the last- and duration of the abnormal electrical dis- mentioned symptom is the most frequent, charge, the symptoms may be: severe stress, still 20% of patients do not display it [3]. nausea, a sinking sensation in the stomach, Parkinson’s disease is often accompanied by a dreamy feeling, visual sensations, unusual dementia. taste, unintelligible speech, confusion, a loud cry (before generalised tonic-clonic seizure), The nursing personnel should initiate ap- tongue-biting, loss of consciousness, inconti- propriate interventions to help the patient nence and apnoea. After the seizure the pa- to cope with these symptoms. An efficient tient may complain of fatigue, headache and communication system should be estab- weakness and may fall into a sleep [4]. lished that will enable patients to adjust communication in accordance with their al- Given this variety of symptoms, the nursing EANM tered cognitive abilities. Emotional support personnel should have good knowledge of should be provided. The patient should be this complex disease so that they can sup- protected from potential injuries by creation port the patient in dealing with, and teach of a safe environment and should be encour- the family how to react during, the epileptic aged to perform physical activities so as to attack. First, the patient should be encour- maintain mobility and physical performance. aged to express his or her feelings about the Physiological support will be provided: the condition; support will be offered to enable patient and the family will be taught about the patient to understand the disease. The the disease and any concerns should be lis- importance of compliance with the pre- tened to and addressed through appropriate scribed medication must be underlined and advice [4]. the safety and efficacy of the medication should be stressed. The drug blood levels Epilepsy should be checked regularly [4]. The symptoms of epilepsy differ mainly due to variations in the origin of pathologi- The family will be instructed to avoid restrain- cal electrical discharges in the brain. There ing the patient during an epileptic seizure, to are two main forms of epilepsy: generalised lie the patient down, to place something soft epilepsy, in which the abnormal electric dis- under the head, to remove hard objects from charge arises in the deep structures of the the vicinity of the patient, to ensure an open brain and spreads throughout the entire cor- airway by turning the patient’s head appro- tex, and focal epilepsy, in which a localised priately, and, if the mouth is open, to place area of the brain is involved. The typical major a soft object between the teeth so as to pro- symptom of an epileptic patient is the tonic- tect the tongue [4]. clonic seizure, formerly known as grand mal

121 Cerebral ischaemia and stroke numbness or weakness, vertigo, ataxia, bilat- Cerebral stroke is the irreversible interruption eral loss of use or feeling of arms and legs, of blood supply to the brain tissue caused by dysphagia or dysarthria, depending on the two main pathological events: occlusion of affected cranial nerve. Finally, small vessel arteries, termed cerebral ischaemia, and ce- ischaemia may produce pure loss of use of rebral haemorrhage, the latter being much or feeling in the contralateral arm and leg [3]. more destructive than the former [3]. When the blood flow to the brain is reduced, isch- Nursing interventions are complex and dif- aemia occurs, which is a reversible condition; fer according to the type and localisation if the condition is severe or prolonged, cell of ischaemia. The airway and oxygenation death ensues, which is termed infarction. If should be maintained and constricting complete recovery from ischaemia occurs clothes should be removed. In unconscious within minutes or hours, the episode is re- patients the lateral position should be ad- ferred to as a transient ischaemic attack. If opted, and mechanical ventilation or oxy- the ischaemic condition lasts more than 24 gen should be provided if necessary. The hours, neurologically one speaks of stroke [3]. vital signs and neurological status should be monitored, including blood pressure, heart The associated symptomatology varies rate, respiratory movements, pain, pupillary greatly depending on the involved arterial changes, sensory function and speech. The branch (and the location and size of the af- nursing personnel must pay attention to fected brain territory) and the length of time the electrolyte and fluid balance and ensure for which the ischaemia persists. The main adequate nutrition (in terms of quantity and neurological symptoms and signs in brain quality). Gastrointestinal activity should also ischaemia are as follows: In middle cerebral be monitored, with attention to the diet in artery ischaemia there may be loss of use of order to ensure efficient digestion and elimi- and feeling in the contralateral face and arm, nation. Adequate hygiene is to be ensured dyslexia, dysphasia, dysgraphia and dyscal- through specific care actions to ensure that culia. Anterior cerebral artery ischaemia pro- the body, eyes and mouth remain clean. duces loss of use of and feeling in the con- The nurse must pay special attention to the tralateral lower limb. Contralateral homony- patient’s position, and this is especially im- mous hemianopia is suggestive of posterior portant in unconscious patients. The body cerebral artery ischaemia. Internal carotid ar- is to be aligned in the correct position, and tery ischaemia involves the face, arms or legs special attention is required to avoid pressure with or without homonymous hemianopia. sores and to prevent pneumonia by chang- Monocular loss of vision indicates ophthal- ing the patient’s position regularly. Patients mic artery occlusion. Vertebrobasilar artery will be taught and encouraged to perform ischaemia produces double vision, facial exercises. Communication with the patient

122 Chapter 10 Health Care in Patients with Neurological Disorders

is to be established and maintained accord- workers, etc. [5]. Effective nuclear medicine ing to the patient’s abilities; the family will practice in relation to neurological diseases be instructed how to communicate with the presupposes efficient interdisciplinary work patient and psychological support will be of- between the referral team, the neurologist, fered [4]. the nursing personnel specialised in neurol- ogy, the nuclear medicine team (in which Multiprofessional aspects of nuclear the nuclear medicine physician will have an medicine practice related to neurological in-depth knowledge of neurology) and other diseases specialist personnel, among whom the tech- Healthcare systems are becoming more nologists have an essential role. Neurological complex owing primarily to scientific and patients are patients with special needs and technical progress, but also because of so- the technologist, as the person in the medi- cial and economic factors related to health cal team who has the most interaction with EANM insurance systems, societal factors such as the patients, should therefore have a strong patients’ expectations and society’s demands knowledge of the field and possess the skills on the medical system. It is impossible for to deal with this category of patient. In re- one individual to cover all necessary aspects cent decades a change towards overlapping of healthcare, especially in the age of ultra- of competencies or responsibilities has been specialisation. Collaborative multidisciplinary noted in medical professions, and this trend work is essential and will provide benefits may be regarded as normal in such a dy- to patients on the basis of the expertise of namic and challenging profession as nuclear different professions, including physicians, medicine [6]. nurses, therapists, technologists, social

123 Figure 1: Schematic r

Assessing - identify priorities for assessing the process - systematic collection of data: nursing history, nursing ndings - continuous updating of database - data validation and communication to medical team Diagnosing - interpretation of collected data - identi cation of health problems - formulation and validation of nursing diagnoses - prioritisation of nursing diagnoses Reassessing Additional data input occurs Deletion or changing of nursing diagnoses Outcome Identifi cation and Planning The established outcomes are not met - the patient/family and the nursing personnel work together to establish priorities - identi cation of expected patient outcomes - establishment of evidence-based nursing Changing the interventions according to nursing plans outcomes Evaluating and in congruence with other therapies - measuring how the patient has achieved - recording and communication of nursing the established outcomes plan - identifying the factors that contributed to the success or failure of the nursing plan - modifying the plan as needed

Implementing - applying the established plan of nursing interventions according to and/or based on available resources, nurses’ abilities, standards of nursing, research ndings and ethical and legal in uences The established outcomes are met - data are continuously collected and the nursing plan is modi ed as needed - all actions and decisions are documented and communicated to the patient/family and to the medical team

The nursing process is CLOSED and the patient has recovered his or her previous state of health

124 5. 4. 2. 1. References References Chapter 10 3. 6.

Taylor CR,LillisC,LeMone P, Lynn P. Fundamentals of Pestean C, Veloso-Jeronimo V, HoggP, eds. Radionu Ellis JR, Hartley CL.Nursingintoday’sEllis JR,Hartley –trends, world Massachusetts: Blackwell Publishing; Blackwell 2005. Massachusetts: I, LennoxWilkinson G, eds. Essential neurology. 4th ed. 2013. of NuclearMedicine; clide metabolictherapy. Vienna: European Association 2011. Wilkins; & WilliamsPhiladelphia:Kluwer/Lippincott Wolters nursing. andscienceofnursingcare. 7thed.The art nuclear medicine. JNuclMed Technol 2011;39:240–8. in andadvanced practice document onentry-level SE,Knapp Dennan W, etal. discussion Euro-American Waterstram-Rich HoggP, K, Testanera H, G,Medvedec 2012. & Wilkins; WilliamsKluwer/Lippincott issues andmanagement.10thed. Philadelphia: Wolters 1994. 2nded.of nursingpractice. Pennsylvania: Springhouse; Buxbaum BS, Mauro E, Norris CG,Buxbaum BS, Mauro E, Norris eds. manual Illustrated - 125

EANM Imprint

ISBN: 978-3-902785-11-4

Publisher: European Association of Nuclear Medicine (EANM) Schmalzhofgasse 26, 1060 Vienna, Austria Phone: +43-(0)1-890 27 44 | Fax: +43-(0)1-890 44 27-9 Email: [email protected] | URL: www.eanm.org

Editors: Fragoso Costa, Pedro (Oldenburg) Santos, Andrea (Lisbon) Vidovič, Borut (Munich)

English Language Editing: Rick Mills

Project Management: Sonja Niederkofler

Content: No responsibility is taken for the correctness of this information. Information as per date of printing September 2015.

Layout & Design kreativ · Mag. Evelyne Sacher-Toporek Linzer Strasse 358a/1/7, 1140 Vienna, Austria Phone: +43-(0)1-416 52 27 | Fax: +43-(0)1-416 85 26 GRafik & desiG n Email: [email protected] | URL: www.kreativ-sacher.at k r e a tiMag. v Evelyne Sacher-Toporek

126 EANM