Tumour Scintigraphy and PET

Tumour cells are de-differentiated from the cell of origin, with loss of contact inhibition, altered cell surface receptors, failure of apoptosis, disordered vessel growth, altered transport systems and unregulated energy use. This causes rapid growth, relative hypoxia and high rate of glucose uptake. γ-emitters are not easily incorporated into molecules without disrupting biological function due to their bulk; hence difficult to have γ-labelled molecules with behaviour identical to target behaviour. Due to their size, PET-emitters can be incorporated into molecules with behaviour identical to original, but have lower half-life. Hypervascularity may increase activity cf background, even when no tumour is present; whereas reduced blood flow (ischaemia, infarction) may reduce activity even when a tumour is present.

Ideal tumour imaging agents:

  • Taken up specifically by the target tissue; if normal tissue uptake occurs it is best that this tissue is uncommonly involved by the tumour.
  • For short-lived isotopes, background activity should be cleared rapidly; longer HL isotopes can be slower (eg I-131 abdominal imaging can be delayed a day to allow secretions to pass).
  • Metabolic trapping in a tumour increases target:background activity (eg Tc99m-MIBI).
  • Ability to predict effectiveness of therapy, eg iodine uptake increases effectiveness of I-131 treatment.

Markers of metabolism include Tc99m-MIBI and Tc99m-TFos (tetrofosmin). These agents are lipophilic, cationic with uptake dependent on cellular and mitochondral membrane potentials. Malignant tumours have negative potentials with high numbers of mitochondria (due to metabolic rate). Hence are indirect markers of metabolic activity (cf FDG being a direct marker). Tumours expressing the multridrug resistence gene P glycoprotine (P-gp, PGP) (in some breast, lung cancers, multiple myeloma) actively transport Tc99m-MIBI and TFos out of the cell causing rapid washout; hence tumours that take up the radioisotope (PGP not present) are likely to be more responsive to chemotherapy. Metabolic agents are otherwise sensitive, but nonspecific. Tc99m-MIBI and TFos for this purpose has been superseded by use of FDG-PET.

Monoclonal antibodies (mAbs) may have some cross reactivity with tissue similar to target. Most utilise the IgG antibody, which consists of 2 variable antigen-binding regions (Fab fragments, contain hypervariable antigen-binding sites) and a constant effector mediate region (Fc fragment, for host binding, also nonspecific binding to liver). Fab fragments are held together by disulfide bonds; Fc bound to Fab fragments with covalent bonds; Fc fragments are able to be removed while maintaining Fab binding (Fab2 fragment). These smaller fragments are exposed to antigen more rapidly, but for a shorter period of time and are cleared from the blood more rapidly; hence fast labelling and rapid background clearance for imaging (esp radioisotopes with short HL); are also less immunogenic and less dose to patient. Larger fragments (intact antibody) used when long reaction time required (weak/slow binding), for treatment (maximised dose to tumour), radioisotope with longer HL; but due to Fc there is nonspecific binding to the liver (may be subtraucted with eg Tc-99m sulfur colloid). The smaller Fab2 fragments increase renal binding; reduced with unlabelled leucine or amino ascites which saturates the renal binding sites. The monoclonal antibodies are created using mouse (murine) hybridoma cells; hence immune response may produce human antimouse antibody (HAMA) causing symptoms; repeat administration may result in fatal anaphylaxis.

Peptides are smaller than mAbs, acting as hormones, neurotransmitters, neuromodulators, growth factors etc; esp brain and GI system. May be regulatory, inhibitory or stimulatory, acting at very low concentrations. Many tumours haver receptors for different peptides.

  • Indium-111-DTPA-octreotide (Ind-111-OCT, Octreo-Scan) binds to somatostatin receptors of neuroendocrine tumours, carcinoid tumours, some neural tumours (meningiomas, medulloblastoma, most astrocytomas, neuroblastomas) and small cell lung carcinomas; also in active granulomas (Wegener granulomatosis, RA, TB). The somatostatin sst2 receptor with ligand is internalised, metabolically trapping the radioisotope. >90% sensitivity for carcinoid tumours, 65% medullary thyroid carcinoma, 90% small cell lung cancer. Rapid renal clearance, 90% by 24hrs.

Ga-67 mimics free iron, binds to transferrin receptors (overexpressed in tumour cells), also taken up by siderophores in pathogens. Initially used for lymphoma and HCC, now superceeded by FDG PET and MR/CT respectively due to high energy γ emitted reducing resolution.

Lymphoscintigraphy used in correlation with sentinel node biopsy. Cutaneous melanoma and breast cancer spread to nearest draining node without skip lesions, hence if this node is negative then extensive lymph node dissection can be avoided. Less useful in melanoma due to poor prognosis despite treatment. Colloid agents include Tc99m-HCA (human serum albumin, smaller size reduces retention in LN, generally better for imaging), Tc99m-SC (sulfur colloid, larger particle size causes retention at injection site, generally better for surgery). For breast, most done as periareolar subdermal injections (other options include intratumoral injection, axillary subdermal). A transmission scan or other body image scan (eg bone scan) used for anatomical localisation. Intra-operatively a gamma probe is used to find the sentinal node.

Other agents:

  • Metaiodobenzylguanide (MIBG) – For neuroblastoma, pheochromocytoma.
  • Iodinated cholesterol derivatives – For adrenal cortical carcinoma.
  • Tc-99m-MIBI – For benign parathyroid tumours.
  • Dual agent to improve detection of metastases in which may have de-differentiated. FDG and I-131 may be used for thyroid cancer. Tc-99m-MIBI and In-111-OCT for multiple drug resistance in small cell lung cancer.

Fusion imaging superimposes nuclear medicine scan to a CT or MRI using fiduciary markers. Hybrid imaging is a single system providing both without patient movement between scans.

Positron Emission Tomography (PET)

Physics of PET.

Scans can be acquired in 2D mode (with septa) or 3D mode (retracted septa)

Most PET-CT is used for oncology, neurology (dementia, epilepsy), cardiac (CAD, myocardial viability) and infection/inflammation (PUO, immunocompromise). In oncology physicians change intended management in 60% of patients after PET, half of which is a change from treatment to non-treatment strategy. Adding PET to conventional work-up will prevent unnecessary surgery in 1/5 of NSCLC.

In general high grade malignancy will demonstrate increased FDG, C-MET, F-DOPA, F-MISO uptake. FET-PET is good for demonstrating extension of tumour.

Radiotracers for use in PET include:

  • F-18 – half-life 110min
    • FDG (F-18 fluorodeoxyglucose) – metabolism
    • NaF – bone imaging
    • F-18 fluoride
    • FLT (F-18 fluorothymidine)
    • FCH (F-18 fluorocholine)
    • FAZA (F-18 fluoroazomycin arabinoside)
    • FET (F-18 fluoroethyl-tyrosine).
  • Ga-68
    • GaTate/DOTA-TATE (Ga-68 DOTA-octreotate)
    • PSMA (Ga-68 prostate-specific membrane antigen) – protein found on membrane in 90% of prostate Ca, is different to PSA. Normal expression in renal tubules, prostate epithelium, duodenum, colon, coeliac ganglia (linear between level of CA and SMA origins, may be assymetrical). May be taken up in other adenocarcinomas esp clear cell RCC, pancreatic.
    • GaNoc/DOTA-NOC (Ga-68 DOTA-[NaI3]-octreotide)
  • I
  • N-13 – half-life 10 min
    • Ammonia – perfusion
  • O-15 – half-life 2 min
    • H2O – perfusion
    • CO – blood volume
    • O2 – metabolism
    • CO2 – blood flow
  • C-11 – half-life 20 min
    • Numerous applications
  • Rb82 – half-life 1min
    • In saline – perfusion

Standardized uptake value (SUV) – Semi-quantitative estimation of lesion activity corrected for attenuation and normalised for injection dose and body weight. SUV (microcuries/cm3) = Activity per unit volume/(Injected Activity/Body Weight). Maximum SUV is a better parameter than average SUV due to heterogeneity of the tumour.

Cancers tend to have high anaerobic metabolism via glycolysis (Warburg effect), with high glucose demand thus up-regulation of cell membrane GLUT-1 transporters and thus high FDG avidity. Most tumours (except HCC) also downgrade glucose-6-phosphatase which reduces exit of glucose out of the cell. This compares with normal cells which have a much higher proportion of aerobic metabolism via pyruvate and mitochondria. Some cancers are not as metabolically active or have differing metabolic pathways (hence not as FDG-avid), but may be more proliferative (eg prostate), which have high cell membrane (phospholipid) turnover during initial growth phase (G1) of the cell cycle, and thus high demand for choline.

Bone scans demonstrate bone reaction to the tumour, rather than the tumour itself. Initially a bone metastasis will be avid on tumour-targeted tracers and cold on bone scan. Only when there is bone reaction to the tumour will bone scans be positive (and hence developing sclerosis), which may continue for a time after the tumour has been eradicated (tends to be peripheral, flare/healing response), later becoming cold.


  • Attenuation correction artifacts – Overcorrection from highly attenuating objects (including contrast), accentuated with movement between PET and CT. Photopenic region on uncorrected PET.
  • Misregistration – Patient movement between PET and CT. Peristalsis may displace bowel contents between the PET and CT, with foci seen adjacent to the bowel.
  • CT truncation artifacts – Objects outside CT field of view, causes dark lines on attenuated-corrected PET. Esp obese patients or arm/leg movement. Algorithm cannot attenuate-correct for objects outside of CT field.
  • Surface contamination of tracer –urinary contamination or leak from injection site. Very difficult to clean off activity.

PET/MR will usually use DIXON VIBE sequences to create a μ map (similar to CT) for attenuation correction.


2-fluoro18-2-deoxy-D-glucose (FDG) is a glucose analogue, actively transported through cell membranes (GLUT1 and less prominent GLUT4 receptors), phosophorylated by hexokinase to FDG-6-phosphate and is unable to be metabolised further; thus trapped and concentrated in metabolically active areas. A completing enzyme to hexokinase is glucose 6-phosphatase which de-phosphorylates FDG, which is then able to exit the cell. The liver is able to clear FDG-6-phosphate. FDG undergoes glomerular filtration, unable to be reabsorbed in proximal convuluted tubules (unlike glucose) hence intense urine activity. Most cancer cells upregulate GLUT1 receptors and downregulate G6P to increase.

Patient fasts (except water) 4-6hrs prior to reduce serum glucose/insulin (hence muscular activity). Blood glucose should be controlled to limit competitive FDG uptake; if high there will be diffusely increased uptake (with normal urinary excretion). High insulin states (eg non-fasting, insulin injection) will result in diffusely increased muscular uptake. Steroids may induce hyperglycemia. Strenuous exercise avoided for 24hrs prior and immediately after injection. Advise plenty of fluids to facilitate urinary FDG clearance, empty bladder (+/- IDUC +/- bladder saline wash +/-loop diuretic). Keep patient warm in hour before injection to limit brown fat uptake (can also be suppressed with benzodiazepine or beta-blocker). Scan 60-90min post-injection (at least 45min) to allow uptake and clearance of background, scan takes 30-60min. In uptake phase should avoid activity, speech, and for brain imaging quite dimly lit room.

Most PET scanners have a bed step/position of approximately 15cm, with 2-5min per step (longer for brain or limited ROI). FOV is usually 70cm in the body, can reduce this to 35cm in the head and neck. When imaging head and neck arms down; for body arms up to reduce beam hardening artefact. Skull base to proximal thigh (EAM to mid thigh) for most studies. Whole body long scan (vertex to toes) for tumours eg high likelihood skin/MSK involvement eg melanoma, T-cell lymphoma). Scanning with CT contrast may cause attenuation correction artifefact, but this is usually minimal, particularly with newer scanners with better AC factors. CT for AC of lungs should be done free-breathing or mid-inspiration; deep inspiration (although better for CT) will cause marked misregistration.

FDG uptake in breasts, but is not secreted into breast milk. Hence recommended to avoid breastfeeding for 12hrs after, can still express and feed.

Quantitative analysis requires sophisticated kinetic studies. Semiquantitative estimation can be performed with SUV or ratio of uptake to internal reference (mediastinal blood pool, liver, cerebellum). Most malignancies have SUV 2.5-3.0, physiologic activity usually 0.5-2.5. SUV dependent on LBW, state of hydration, insulin level, blood sugar level, distribution of nontarget organs. A change in SUV by >20% between scans is a probable true change (if uptake time and blood glucose similar).

Oncological indications include NSCLC, SPN, lymphoma, sarcomas, recurrent GI carcinoma (colorectal, gastro-oesophageal, pancreatic), metastatic melanoma, H&N cancer, cervical/ovarian carcinoma, thyroid carcinoma. Intensely avid tumours (SUV 7-16) include HG lymphoma, colon, NSCLC, oesophageal, H&N, HG sarcoma, melanoma. For initial staging, evaluation of response and assessment of recurrence. False negatives include tumours with low activity, lesions <2x scanner resolution (<10mm), tumour necrosis, recent chemo/RT/high-dose steroids, hypergycaemia. Tumours with low uptake include lobular breast, necrotic/mucinous tumours, AIS/BAC, well-differentiated HCC, RCC, prostate, neuroendocrine/carcinoid, thyroid; low grade lymphoma, sarcoma, sclerotic bone metastases. False positive uptake includes granulomatous lesions (sarcoidosis, TB), infection (fungus, bacteria, mycobateria), phagocytic activity in necrotic tissue (<2-4 weeks post-chemotherapy), bone marrow reactivation, thymic rebound, hypermetabolic brown fat.

  • Brain – Activity correlates with tumour grade, except pilocytic astrocytomas (low-grade but avid). Activity >1.5x white matter or 0.6x normal grey matter suggests malignancy. Slice-by-slice correlation with anatomy is required. Can be used to plan biopsy, targeting high grade regions. Post-op there should be no appreciable activity, even early. Radiation induces inflammation, which is avid in acute phases, hence need to wait 2-4/12 before scan. Areas of high activity may represent active seizure centres. Brain lesions only reliably seen when >20mm.
  • Salivary glands normal minimal-low uptake, increased with radiotherapy, infection, inflammation. Activity in salivary secretions retained in gingival recesses, vallecular etc, improved with rinsing mouth with water. Normal variable uptake in Waldeyer’s ring; focal areas of uptake should undergo visual inspection.
  • Head and Neck – Better than CT and MR for nodal disease. Reliable occult malignancy detection. 15% have other malignancies (lung, oesophageal) which can be detected with PET. Unilateral laryngeal uptake may suggest vocal cord paralysis. Activity in dental disease, tracheotomy sites, postradiation mucositis, oesophagitis, gingivitis. Benign tumours may be avid (eg Warthin tumour). Post-treatment PET for recurrence best delayed 2-3months.
  • Thyroid – Diffuse low level. Diffusely increased uptake in Hashimoto > Graves thyroiditis; recommend correlation with TFTs.
  • Lung cancer – Most useful in non-small cell cancers, most of which are highly avid. Hypometabolic malignancies include AIS/BAC and carcinoid. Sites of tumour involvement seen in 11% when not suspected on CT. CT may overestimate primary lesion size due to adjacent consolidation/atelectasis, improved with PET. PET reduces the number of mediastinoscopies required for staging. Activity in pleural effusion suggests malignancy, unless post-pleurodesis; but absence of activity doesn’t exclude involvement. Radiation pneumonitis causes uptake in 1st 6/12. Reduction in SUV >75% suggestive of CR; >25% PR; residual tumour may need definitive chemoradiation. Restaging is most useful in stage III disease which has high risk of recurrence in 3-15 months.
  • Solitary pulmonary nodule (SPN) – Positive if activity greater than mediastinal blood pool, but any uptake if nodule <10mm is suspicious. Generally needs to be greater than 8-10m in size for accurate assessment. SUV may be artificially low due to smearing artifact from breathing (improved with respiratory gating), size <20mm (partial voluming). Limited by low FDG-avid tumours (eg carcinoid, adenocarcinoma in-situ), ground-glass nodules, near diaphragms (movement). Radiographcis 2014 ?2012 PET nodule.
  • Blood pool – Can be used as a reference. Mediastinal blood pool SUV taken from ascending aorta or arch.
  • Thymus – uptake in reactive thymic hyperplasia ranging from 1 to >4 months post chemo; triangular in children, patchy in adults.
  • Heart – Esp after eating as fasting switches to fatty acid metabolism. Even if fasted uptake is variable and nonuniform esp base of LV. If need to suppress uptake will need to fast for ~12hrs. Ischaemia will increased FDG uptake.
  • Breast – Reliable for lesions >10 mm, but poor sensitivity for small lesions, DCIS, LCIS or low grade lobular carcinoma. PET is not recommended for staging of early stage disease (I-II) or axilla (low sensitivity); SLN biopsy much best for this. Useful to evaluate brachial plexus if other imaging studies normal. May be helpful for monitoring, as high accuracy for detecting response to treatment (SUV <55% baseline), particularly mid-course, as some nonresponding tumours will show a reduction in uptake later in treatment (?altered metabolic pathway).
  • Oesophagus – All subtypes have high activity, but low (50%) sensitivity due to small size, high specificity. Can detect lesions >3-5mm. Pre-scanning drinking with water clears activity from salivary secretions. Difficult in distal oesophagus due to oesophagitis, GOR, Barrett’s oesophagus, hiatus hernia and retained saliva. Limited sensitivity of nodal metastases. High sensitivity but low specificity for tumour bed recurrence. High sensitivity and specificity for distant metastases.
  • Stomach – Limited assesssment of primary due to normal mucosal activity. Nodal disease more accurate. Hepatic disease limited.
  • Pancreas – PET lacks resolution for vascular invasion, but can detect distant metastases. Gastric activity limits pancreatic bed assessment, improved with water intake. Useful for discriminating cancer from chronic pancreatitis.
  • Liver – Low level activity in almost all, usually appears slightly heteogeneous. Normal SUV average 2-3, max 3-4. High metabolic activity and variable clearance from hepatocytes. Usually used as a reference, with SUV taken from posterior right lobe. PET almost always negative in benign lesions; positive lesions mostly malignant or abscess/hepatitis. However some malignant lesions can be non-avid, eg HCC variable uptake depending on grade. HCC differentiation correlates with uptake. Not useful for HCC diagnosis due to poor specificity as non-malignant lesions eg adenomas may be avid. Cholangiocarcinoma seen if nodular/focal, poorly seen if infiltrative. Inflammation from biliary obstruction is avid. Metastases <10mm not well seen due to resolution and background activity.
  • Spleen – Normal uptake more than blood pool, less than liver. Increased by haematopoiesis.
  • Gallbladder – Cancer is highly avid and well seen. Pericholecystic activity is usually inflammation.
  • Kidneys – High renal collecting system activity, reduced with good hydration. Limited primary detection due to normal parenchymal activity. Very accurate for extrarenal involvement.
  • Ureter and Bladder – Limited due to urine activity. Ureteric activity may be focal due to peristalsis, hold up in areas esp pelvic brim. Bladder diverticula may retain urine after bladder emptying. Useful for local nodes and metastases.
  • Adrenals – abnormal if activity more than liver. Adenomas are usually <liver; rarely >liver in which case can still be diagnosed as adenomas if <10HU.
  • Colorectal – Normal uptake is highly variable, usually > SB, esp caecum, descending and sigmoid colon. Activity from peristalsis, metformin, mucosal inflammation, lymphoid tissue in wall, colonic bacteria. Multifocal/segmental in IBD. Intense focal uptake is seen in 3%, at least half of which is malignant/pre-malignant including hyperplastic polyp/adenoma, carcinoma; also seen with peritoneal implants or LN metastases. Diverticula, polyps and faecal material may accumulate FDG.
    PET not routinely indicated for routine staging unless initial studies suggestive but not conclusive for metastatic disease, or if considering metastectomy. Pre-operative PET associated with reduction in recurrence rate and better 5-yr survival. Low sensitivity of nodal disease, as usually involved by few tumour cells, and are close to bowel activity. Sensitivity PET better than CT for all metastases (including liver), except lung where they are similar; specificity better than CT in all areas except abdomen. PET not recommended for restaging, unless curative resection is considered. Granulation tissue may be avid, but PET useful for recurrences 6/12 after surgery or radiotherapy. If baseline scan available the PET can assess at 2/12 by comparing original site of uptake with more diffuse uptake of inflammation. Detection of local recurrence more accurate than CT (95% vs 65%), and is complimentary to CEA (elevated in 2/3 of those with suspected recurrence). PET identifies tumour in 84% of patients of rising CEA when conventional working (including CT) is negative.
  • Cervix – Very useful for local nodes, response, recurrence.
  • Uterus – Intense endometrial and ovary/Fallopian tube uptake during menstruation and ovulation. Activity in menstrual blood occasionally in vaginal tampons. Pelvic endometriosis shows high activity. Useful for local nodes, peritoneum, distance metastases, recurrence. Subtle peritoneal seeding requires slice-by-slice CT correlation. Limited by peri-ovulatory and fibroid activity.
  • Ovaries – Uptake in functional cysts. Small peritoneal metastases <5mm usually undetected. Peri-ovulatory follicles can be avid.
  • Prostate – BPH usually not high uptake. Focal uptake may be inflammatory or carcinoma (which may not necessarily have high PSA ?due to de-differentiation). Well-differentiated cancers tend to demonstrate sclerotic bone metastases, hot on bone scan. Less-differentiated tumours demonstrate increasing FDG uptake, with lytic bony metastases. NaF PET cf Tc-MDP bone scan is more sensitive and specific, faster, bettern target-to-background. FCH also detects extra-osseous lesions; for bone mets is less sensitive, but more specific than NaF. PSMA is very sensitive and specific, essentially replacing NaF and FCH in prostate cancer, with much higher target-to-background than FCH, may detect lesions as small as a few mm. PSMA is expressed higher in higher grade tumours. FDG remains useful to detect aggressive de-differentiated disease. Radionuclide therapy using PSMA in research phase.
  • Testicles – Very useful for metastases. Seminoma usually more avid than non-seminoma.
  • Lymph nodes – High sensitivity, specificity and negative predictive value. False positives in sarcoidosis, TB, pyogenic abscess, histoplasmosis, fungi, discitis. Hypermetabolic nodes are considered malignant until proven benign (biopsy).
  • Lymphoma – Uptake correlates with histological grade. Low grade lymphomas (esp follicular NHL, MALTomas) have limited uptake. When CLL involves nodes = small cell lymphoma, usually low grade uptake; SUV >5 is highly suggestive of Richter’s transformation. PET may help to upstage disease, but rarely does this change treatment. However PET is useful as a baseline for monitoring treatment, and can help to target biopsy to most avid disease. ½ of untreated cases with diffusely increased marrow uptake is hyperplasia, ½ diffuse lymphoma infiltration. Diffuse post-chemo uptake is hyperplasia. Heterogeneous/nodular increased uptake is lymphoma, can be confirmed with targeted biopsy.
    A drop in uptake after chemotherapy is detected earlier than CT nodal size, with 80-90% of SUV drop occurring in the first 7 days. A negative PET in presence of residual mass is still interpreted as complete response. Rresidual activity may be tumour or inflammation with ongoing necrosis. In NHL if there is residual activity standard is to continue same chemo treatment. In HL if there is residual disease (>blood pool/liver) then may benefit from change in treatment. CMR if </= mediastinal blood pool (or less commonly used liver). Douville grading. New pulmonary lesions in the setting of improved/resolved lymphoma almost always inflammatory. Some chemotherapy drugs (esp Bleomycin) may cause pneumonitis with diffuse uptake. Granulocyte colony-stimulating factor may cause focal marrow activity. Healing from radiotherapy or surgery shows uptake. Thymic activity from rebound is common in children undergoing treatment.
  • Bone marrow – Mild-moderate, diffuse in spine, pelvis, ribs, sternum, proximal femurs. Normally similar to spleen, less than liver, more than blood pool. Increased uptake with marrow activation (haematopoiesis eg GCSF), marrow rebound, anaemia, chronic disease/infection. GCSF in many chemo regimes, made by some tumours. Reduced uptake in areas of radiotherapy (fatty marrow change).
  • Melanoma – Primary tumour is hot, but due to small volume tumour in nodes is limited (very low sensitivity, but high specificity), hence SLN biopsy should be done. Acne and carbuncles may show activity. PET very useful for distant metastases with high sensitivity and specificity (apart from brain).
  • Musculoskeletal – Skeletal muscle uptake negligible in resting, increased with heavy activity in preceding 24hrs. SOB may cause chest wall and diaphragmatic uptake. Increased uptake in insulin injection. Eye movement causes ocular muscle uptake. Talking causes uptake in larynx and muscles of mastication. Arthritis and osteophytes may be avid. Fractures demonstrate activity for weeks from haematoma resorption, granulation tissue, callous formation. Useful for staging, nodes, metastases. High sensitivity and specificity for non-sclerotic bone metastases (sclerotic mets typically less aggressive).
  • Injection leakage – At injection site, draining lymph nodes or indwelling cather tip if thrombus present.
  • Brown fat – Symmetric uptake in paraspinal, mediastinum, neck, supraclavicular regions with corresponding fat signal on CT. Increased activity when patient is cold or anxious (increased catecholamines increasing glycolytic activity), reduced with warm rooms and sedation.
  • Prostheses – true periprosthetic uptake is nonspecific, usually of no significance, can last forever (?due to macrophages).

Physiological uptake in brain, salivary glands, lymphoid tissue thyroid, brown fat, thymus, lactatic breast, areola, muscles, GIT, urinary tract, female genital tract.

Benign lesions with uptake – pituitary adenoma, adrenal adenoma, thyroid follicular adenoma, salivary gland tumours, colonic polyps, ovarian thecoma/cystadenoma, GCT, ABC, leiomyoma, marrow/splenic hyperplasia, thyroiditis, Cushing’s, thymic hyperplasia, fibrous dysplasia, Paget’s disease, hibernating myocardium, other inflammatory processes.

Post-radiation change esp in liver, lung may be focal. Common flare in inflammation esp 6-9 months post XRT. FDG-avid radiation liver injury is reversible, may see associated fatty change on CT. Marrow uptake reduced within the radiation field, with fatty infiltration.

For monitoring therapy, activity of the lesions can be summarized as progressive metabolic disease (PMD), stable metabolic disease (SMD), partial metabolic response (PMR), complete metabolic response (CMR). European organization for research and treatment of cancer criteria. CMR suggests effective treatment or tumour cells that have changed metabolism, or quiescent tumour cells which are now in G0 of cell-cycle.


Neurological indications include dementia and epilepsy. Cerebral blood perfusion (CBF) maps can be obtained from O-15 H20 PET. CBF is very similar to FDG uptake due to correlation between blood flow and glucose uptake. Highest activity in cortical grey matter (esp posterior cingulate gyrus, primary visual cortex medial occipital lobe), putamen, head of caudate, thalamus, cerebellum. Apparent reduction in uptake in medial temporal lobes due to partial voluming. FDG is highly dependent on blood glucose due to competitive uptake. Injection in a calm dark room for homogeneous distribution across cortex.

Progressive dementias usually spare the sensorimotor cortex, which would be avid compared to adjacent areas. If advanced, AD will be difficult to distinguish from FTD.

Lewy body dementias – visual hallucinations typical. Hypometabolism similar pattern to Alzheimer’s but more prominent occipital lobe involvement.

Primary progressive aphasia – unilateral (L>R), involving hemisphere, basal ganglia, thalami, crossed cerebellar diaschesis. May involve sensorimotor cortex (cf sparing in AD/LBD).

Depression and hypothyroidism may mimic Alzheimer’s, but will improve with treatment.

Congenital abnormalities do not cause crossed cerebellar diaschesis, as afferent inputs from other regions would have been developed.

Epilepsy – SPECT tracers complete uptake and trapping in minutes of injection hence are ideal for ictal studies. FDG uptake continues for >1hr, hence increased uptake with propagate into areas adjacent to seizure focus. Chronic seizure medications reduce cerebellar activity (may be assoc with atrophy).


FDG-PET is gold-standard for myocardial viability and hibernating myocardium, more accurate than SPECT.

  • Myocardial perfusion uses N-13 NH3, Rb-82 chloride or O-15 H2O. Ammonia rapidly diffuses into myocardium, retained by glutamate pathway with accumulation depending on regional perfusion. Rb-82 is potassium analogue, actively taken up by myocardium, but positrons are more penetrating hence lower resolution. Stress performed with inotropes (dobutamine), vasodilators (dipyridamole) or pacing; with comparisons between rest and peak stress. Ischaemic myocardium has transient reversible stress-induced perfusion defect. Infarction is fixed, nonreversible defect. Hibernating myocardium has poor contractility due to low perfusion, responds to revascularisation; perfusion (PET or SPECT) is compared with metabolism imaging (FDG-PET) for diagnosis. Stunned myocardium has contractile dysfunction despite normal perfusion.
  • Myocardial metabolism uses FDG. Glucose is 2nd energy source after fatty acids. Fasting shifts metabolism to fatty acids, pretreatment with 50g glucose increases FDG uptake. Diabetes causes heterogeneous, nonspecific defects, requires oral glucose loading with continuous infusion of insulin, glucose and potassium to enhance uptake. Increased uptake in ischaemia, absent/reduced in infarction. Areas with perfusion defects with high metabolism (perfusion-metabolism mismatch) indicates hibernating myocardium and benefit from revascularisation. Perfusion-metabolism match indicates nonviable scar tissue. However, acute evolving MI may show high FDG uptake, hence best to image when patient is stable.

Inflammation and infection

  • PUO – PET useful for detecting indolent infection or occult neoplasm as a cause.
  • Immunocompromise (esp AIDS) – For undetected infection or tumour. CNS lymphoma is hypermetabolic compared with toxoplasmosis (no/little uptake).
  • Sarcoidosis – Uptake correlates with activity (treatment indicated), extent of disease and response to treatment.
  • Vasculitis – Uptake in affected vessels with negative predictive value of 80%.


Amyloid Scans

Tracers (eg florabetir) nonspecifically bind to white matter. Cerebellum can be used as a reference. Positive scan (indicating Alzheimer’s) when there is reduced grey-white contrast, ie activity extends into grey matter, usually diffusely.


  • Delbeke, Coleman. Procedure Guideline for Tumour Imaging with 18F-FDG PET/CT 1.0.
  • NCCN task force report: PET/CT scanning in cancer JNCCN Vol5 Suppl1