PET imaging of inflammation and infection
Inflammation is a protective response of living tissue toward injury, caused for example by microbial invasion (infection), trauma, or tumour. Tightly regulated inflammatory processes are also an intrinsic part of normal tissue remodelling and growth. During the inflammatory response the blood flow and capillary permeability are increased and leukocytes migrate from blood into the interstitial space and lymph in the affected tissue region. Acute inflammation is initiated by resident immune cells and platelets, which release inflammatory mediators such as histamine, heparin, and serotonin. Activated tissue macrophages and damaged cells release cytokines which attract and activate the leukocytes, and induce the activation of VAP-1 and expression of selectins and integrin ligands on endothelial cell surfaces and in extracellular matrix. In chronic inflammation the extracellular matrix is degraded by matrix metalloproteases, but also healing processes such as angiogenesis take place at the same time. Activated immune cells may also induce local hypoxia and expression of hypoxia-inducible factors.
Ecto-phosphatase proteins, including alkaline phosphatases, dephosphorylate extracellular inflammation triggering moieties, including lipopolysaccharides and nucleotide phosphates. Conversion of ATP, ADP, and AMP to adenosine leads to anti-inflammatory effect via adenosine receptors (Chandrupatla et al., 2018).
The location of inflamed tissue can be found using PET imaging, which is sufficient in many clinical situations, but it would be useful to distinguish between inflammation caused by normal wound heeling and post-surgical infection, and between sterile and infectious loosening of joint replacements. At present the role of PET in the discrimination between infection and non-microbial inflammation, let alone identification of pathogens, is limited, but developing rapidly.
Numerous inflammation-specific PET radiopharmaceuticals have been introduced. Many radiopharmaceuticals that have been aimed for tumour imaging have been found to accumulate to the tumours at least partly because of local inflammation.
Inflammasomes consist of heterogeneous group of cytosolic multi-protein complexes, forming the key components of the innate immune system, connecting to the adaptive immune system (Deets & Vance, 2021). Each inflammasome contains a specific sensor protein. Detection of certain danger signals leads to activation of caspase proteases, which further lead to release of pro-inflammatory cytokines such as IL-1β and IL-18. PET radioligands for inflammasome imaging are being developed (Xu et al., 2021).
Numerous techniques have been developed for labelling white blood cells (WBC) with isotopes for tracing the sites where WBCs accumulate. For SPECT imaging, lipophilic compounds [99mTc]HMPAO and [111In]oxyquinoline are used to label autologous leukocytes (Love & Pellegrino, 2004). For PET imaging, [18F]FDG-labelled white blood cells are the most studied (Pellegrino et al., 2005). [89Zr]oxinate4 could be used for long-term tracking of leukocytes (Charoenphun et al., 2015). Usually, though, white blood cells are labelled in vivo, by administering radiopharmaceutical that binds to targets that are abundant on white blood cells, such as TSPO, chemokine receptors, formyl peptide receptor, cellular adhesion molecules such as vascular adhesion protein 1, folate receptors, CB2 receptors, somatostatin receptors, and P2X7 receptors, or systems that are over-active in activated WBCs, such as cholinergic system, deoxyribonucleotide salvage pathway, and glycolysis. Activated granulocytes release enzymes, such as neutrophil elastase, which can be detected with PET radioligands.
Activated inflammatory cells have an increased demand for glucose. Increased expression of glucose transporters and hexokinase results in increased uptake of [18F]FDG in the tissue crowded by activated macrophages and neutrophils. Although increased [18F]FDG uptake is not specific for inflammation, it has been used to study for example rheumatoid arthritis, Lyme arthritis (Pietikäinen et al., 2017), inflammatory bowel disease, sarcoidosis, idiopathic retroperitoneal fibrosis (Vaglio and Maritati, 2016), vasculitis (Taimen et al., 2019), vascular inflammation in atherosclerosis, renal inflammation and infections (Wan et al., 2018, bone and prostheses infections, diabetic foot (Basu et al, 2012), eosinophilic inflammation in asthma, and muscular inflammation (Yamada et al., 1995; Tatejama et al., 2015; Aro et al., 2017). Since [18F]FDG PET can detect both inflammation and cancer, it is optimal for imaging patients with fever of unknown origin (Bleeker-Rovers et al., 2009).
Increased tissue uptake of PET tracers for translocator protein, TSPO, can be observed in regions where macrophages are present, making them more specific inflammation markers than [18F]FDG. TSPO targeted PET tracers are mainly developed and applied for detecting neuroinflammation, for example in neurodegenerative disorders, stroke, and traumatic brain injury. TSPO ligands have also been used to study atherosclerosis, rheumatoid arthritis, and muscular inflammation. However, the relatively high expression of TSPO in normal peripheral tissues may limit the applicability of TSPO imaging outside of the nervous system (Largeau et al., 2017).
Folate receptor β is expressed on activated macrophages, but not in quiescent macrophages. Folate receptor targeted PET radiopharmaceuticals have shown better or equal signal-to-noise ratio, but much lower target-to-background ratio, than [18F]FDG in animal model of inflammation (Kularatne et al., 2013). Because of their better sensitivity to inflammation than [18F]FDG, FRβ-specific radiopharmaceuticals can be useful in drug research and diagnosis of inflammatory diseases. For example, [18F]fluoro-PEG-folate can be used to monitor anti-folate therapy in rat model of arthritis (Chandrupatla et al., 2017). Folate-conjugated anti-inflammatory drugs and FRβ-targeted antibodies are under development for autoimmune diseases.
Activated immune cells may cause a localized tissue hypoxia and expression of hypoxia-inducible factors. PET radiopharmaceuticals developed for hypoxia imaging have been used in experimental rheumatoid arthritis model in mice (Fuchs et al., 2017).
Perfusion and oxygen extraction fraction (OEF) have been studied in rabbit muscle infected with Escherichia coli using steady-state [15O]CO2 and [15O]O2; large increase in perfusion was seen, and moderate decrease in OEF (Senda et al., 1992).
In chronic inflammation, tissue-infiltrated lymphocytes release angiogenic cytokines and chemokines. In a sterile muscle inflammation model, increased uptake of integrin αvβ3 tracer [18F]Alfatide-II was observed (Wu et al., 2014).
Vascular leakage and oedema can be visualized using labelled macromolecules which normally can not leave the vasculature. For example, dextran, transferrin, and albumin have been labelled with 68Ga. Albumin has also been labelled with copper isotopes, and with 18F. Increased vascular permeability in animal model of muscular inflammation has been detected with 18F-labelled albumin (Niu et al., 2014). Specific in vivo labelling of albumin is possible with radioligands containing albumin-binding domain, for example [18F]AlF-NEB (Niu et al., 2014; Wang et al., 2015), [68Ga]NEB (Zhang et al., 2015), and [68Ga]ABY-028 (Jussing et al., 2020). 68Ga3+ injected into body as chloride or citrate also binds quickly to transferrin and albumin.
Gallium-67 was the first radionuclide used for imaging inflammation. 68Ga3+ has shown some promise in imaging bacterial bone infection in animal model (Mäkinen et al., 2005) and in human patients (Nanni et al., 2010; Salomäki et al., 2017). In bacterial infection of soft tissue [18F]FDG resulted in higher signal than [68Ga]Ga3+ (Salomäki et al., 2017). [68Ga]Ga3+ may be useful in detecting infected lower limb prostheses (Tseng et al., 2019).
Apoptosis is often linked to inflammation and infection, and can therefore often be studied using PET radiopharmaceuticals aimed for cell death imaging, even in case of aseptic inflammation (Liang et al., 2014). Additionally, many pathogens can avoid inflammatory response by apoptotic mimicry, which may then be best detected by apoptosis imaging.
[18F]F- PET has shown promise in imaging of increased bone turnover during inflammation of the bone and in rheumatoid arthritis and ankylosing spondylitis. [18F]F- has also been used to image microcalcification in arterial plaques, both in animal disease models and in humans. Increased [18F]F- uptake has also been observed in chronic tuberculosis lesions in mice.
Sphingosine-1-phosphate (S1P) and its receptors are involved in many immune system mediated diseases, including MS, rheumatoid arthritis, and inflammatory bowel disease. PET radiopharmaceuticals targeting S1P receptor 1 have been used in animals models of inflammatory diseases.
Cyclooxygenase (COX), or officially prostaglandin-endoperoxide synthase (PTGS) is a family of enzymes that are responsible for the formation of prostanoids, including prostaglandins and thromboxane. These cause the pain and tissue swelling in inflammation. Prostaglandins are derived from arachidonic acid, which is an abundant fatty acid in cell membranes. Nonsteroidal anti-inflammatory drugs such as aspirin and ibuprofen are COX inhibitors. COX-1 is expressed constitutively in normal tissues, but the expression of COX-2 is induced by several cytokines, including TNF-α.
Several COX inhibitors have been labelled for use in PET (Takashima-Hirano et al., 2010). As an example, [11C]ketoprofen binds to both COX-1 and COX-2, and has been used to observe the status of inflammation, for example in rheumatoid arthritis. COX-1 targeting [11C]PS13 may be used as marker of neuroinflammation (Kim et al., 2020). COX-2 specific drugs have been developed, but the PET radiopharmaceuticals based on those have mostly provided disappointing results (Laube et al., 2013).
Soluble epoxide hydrolase (sEH) degrades epoxides of fatty acids, including epoxyeicosatrienoic acid isomers (EETs) which reduce vasoconstriction and inflammation. sEH inhibitors can be used in treatment of cardiovascular and renal disorders, and for peripheral analgesia. [18F]FNDP can be used to assess sEH activity in the brain (Coughlin et al., 2021).
Neurokinin 1 receptors and substance P
Increased NK1 receptor availability is observed with [11C]GR205171 PET in neurogenic inflammation.
Mannose receptor (CD206) is expressed on the surface of macrophages, dermal fibroblasts, keratinocytes, and mesangial cells. Mannans contain mannose, and are an important part of cell walls of yeasts and fungi. Mannose receptor can be targeted with mannose-containing macromolecules, for instance [18F]fluoromannan and [68Ga]IRDye800-tilmanocept. Locke et al (2012) used mannose coated labelled liposomes in a mouse model.
Fibroblast activation protein (FAP) is expressed on active fibroblasts. Small molecule FAP inhibitors (FAPIs) are internalized after binding, offering good target-to-background ratio in diagnostic imaging and theranostic use.
Transient receptor potential channels (TRPs) support inflammation via secretion of proinflammatory neuropeptides, and induce the sensation of pain. Function of TRPs is modulated by endocannabinoids. TRPV1 and TRPV6 are targets for capsaicin, that can induce apoptosis in some tumour cell lines.
Several PET radiotracers have been developed for imaging bacteria (Auletta et al., 2019), viruses, and other pathogens.
Bacterial cell walls contain molecular structures that can be targeted by specific antibodies, antibody fragments, or affibodies. β-D-glucans on some bacteria and fungi, and lipoteichoic acid (LTA) on Gram-positive bacteria are examples of such targets.
[18F]FIAU was initially developed for imaging of cells transfected with reporter gene (HSV1 thymidine kinase). Cells which express the enzyme will phosphorylate the radiopharmaceutical, leading to trapping of 18F-label. However, [18F]FIAU is also substrate for the endogenous thymidine kinase in bacteria, enabling in vivo visualization of bacteria and the effect of antimicrobial therapy.
Ciprofloxacin inhibits bacterial DNA gyrase and topoisomerases, and the uptake of 99mTc- and 18F-labelled ciprofloxacin has been used for visualisation of lesions with bacterial infection. However, nonspecific binding of [18F]ciprofloxacin is relatively high and it is also retained in granulocytes. Its initial uptake is affected by general inflammatory responses such as increased perfusion and vascular leakage.
Other labelled antibiotics include depsidomycin derivative [68Ga]DOTA-TBIA101, [18F]fleroxacin, 2-[18F]-INH, and [18F]trovafloxacin.
Antimicrobial peptide (AMP), or host defence peptide, is a general term for over thousand molecules identified so far that could be described as natural microbicides, since they are produced by eukaryotic cells, also in humans, to target prokaryotic cells. AMPs are electrostatically attracted to the negatively charged bacterial cell walls, destroying the membranes or inhibiting intracellular processes. Although bacterial mass in infected tissues is small, the large surface-to-mass ratio of bacteria enables imaging applications.
Ubiquicidin and its fragments have been radiolabelled, and shown promise in animal models in PET and SPECT imaging. [68Ga]Ubiquicidin did not accumulate in osteomyelitis lesions in porcine model (Afzelius et al., 2020).
Siderophore ("iron carrier") is a general term for small molecules secreted by most micro-organisms for Fe3+ acquisition and storage (Petrik et al., 2017). Most fungi and bacteria can utilize siderophores released by other micro-organism, even if they themselves have lost the ability to produce siderophores. Since the chemistry of Ga3+ is very similar to Fe3+, 68Ga-labeled siderophores and [68Ga]Ga-citrate have been used to target bacterial and especially fungal infections. Siderophores such as [68Ga]TAFC, [68Ga]FOXE, [68Ga]DFO-B, and [68Ga]PVD-PAO1 could be used as specific infection imaging, in comparison to non-specific 68Ga-citrate or 68Ga-colloids (Petrik et al., 2015, 2018, 2020, and 2021; Misslinger et al., 2021).
Sorbitol is a metabolic substrate for certain strains of gram-negative bacteria. 2-[18F]-fluorodeoxysorbitol ([18F]FDS) has been shown in mice studies to specifically accumulate in tissues with active infection, but not in inflamed tissue (Li et al., 2008; Weinstein et al., 2014; Yao et al., 2016; Ordonez et al., 2017; Li et al., 2018). [18F]FDS can be used to assess BBB disruption induced by focused ultrasound (Hugon et al., 2021).
[18F]FDS is easy to produce from [18F]FDG by a one-step reduction, it is metabolically stable, and rapidly excreted to urine via kidneys (Wakabayashi et al., 2016; Werner et al., 2019). [18F]FDS is suitable for human studies from a radiation dosimetry perspective (Zhu et al., 2016).
Antiviral drugs could be labelled and used in virus-specific PET and SPECT imaging (Bray et al., 2010). [18F]FPMPA is a 18F-labelled analogue of antiretroviral tenofovir.
Many intracellular pathogens exploit the hosts response to apoptosis by exposing phosphatidylserine (PS) on their surface or by cloaking themselves in host cell derived PS-containing vesicles, to facilitate binding, entry, and immune response evasion (Rice et al., 2016). Apoptotic mimicry is utilized by many viruses, such as hepatitis B, HIV, and dengue; and some parasitic micro-organisms, such as trypanosomatids. Also certain bacteria, such as Listeria monocytogenes, emerge from infected macrophages packaged in PS-coated vesicles, thus gaining access to healthy macrophages via phagocytosis.
Although the exposed phosphatidylserine is the natural signal to macrophages to start phagocytosis, also other anionic phospholipids can trigger the same response. Anionic surface charge is a common feature of the bacterial cell envelope, achieved by phosphate-containing lipopolysaccharides (LPS), techoic acids, or phospholipids such as phosphatidylglycerol (PG) and cardiolipin (CL). Escherichia coli, Staphylococcus aureus, and Streptococcus pyogenes are examples of these bacteria.
PET radiopharmaceuticals developed to target PS may be useful on detecting these pathogens, although the response will not be specific, as PS is exposed also in normal apoptosis, even during non-microbial inflammation.
Certain bacteria can produce extracellular polymeric substance, and stay embedded in this matrix which provides them protection from the immune system and antibiotics (Otto, 2018), and may prevent the binding of imaging agents targeted to bacterial surface structures. These "biofilms" aggregate on surfaces such as catheters, artificial heart valves, joint prostheses, and bone. Life-threatening sepsis is often caused by Staphylococcus aureus biofilms.
Phage-display selected peptides with an affinity towards S aureus biofilm have been labeled with 68Ga (Nielsen et al., 2016). In pig osteomyelitis model, no uptake was seen in the lesions (Afzelius et al., 2020).
- Leukocytes and platelets
- Lymphatic system
- Somatostatin receptors
- Endocannabinoid system
- Histaminergic system
- Endothelial barrier
Ady J, Fong Y. Imaging for infection: from visualization of inflammation to visualization of microbes. Surg Infect. 2014; 15(6): 700-707. doi: 10.1089/sur.2014.029.
Auletta S, Varani M, Horvat R, Galli F, Signore A, Hess S. PET radiopharmaceuticals for specific bacteria imaging: a systematic review. J Clin Med. 2019; 8(2): 197. doi: 10.3390/jcm8020197.
Autio A, Jalkanen S, Roivainen A. Nuclear imaging of inflammation: homing-associated molecules as targets. EJNMMI Res. 2013; 3(1):1. doi: 10.1186/2191-219X-3-1.
Basu S, Ranade R. 18-Fluoro-deoxyglucose-PET/computed tomography in infection and aseptic inflammatory disorders: value to patient management. PET Clin. 2015; 10(3): 431-439. doi: 10.1016/j.cpet.2015.03.006.
Blockmans D. PET in vasculitis. Ann N Y Acad Sci. 2011; 1228: 64-70. doi: 10.1111/j.1749-6632.2011.06021.x.
Bray M, Di Mascio M, de Kok-Mercado F, Mollura DJ, Jagoda E. Radiolabeled antiviral drugs and antibodies as virus-specific imaging probes. Antiviral Res. 2010; 88(2): 129-142. doi: 10.1016/j.antiviral.2010.08.005.
Chandler MB, Zeddun SM, Borum ML. The role of positron emission tomography in the evaluation of inflammatory bowel disease. Ann N Y Acad Sci. 2011; 1228: 59-63. doi: 10.1111/j.1749-6632.2011.06032.x.
Dorward DA, Lucas CD, Rossi AG, Haslett C, Dhaliwal K. Imaging inflammation: molecular strategies to visualize key components of the inflammatory cascade, from initiation to resolution. Pharmacol Ther. 2012; 135(2): 182-199. doi: 10.1016/j.pharmthera.2012.05.006.
Glaudemans AW, de Vries EF, Galli F, Dierckx RA, Slart RH, Signore A. The use of 18F-FDG-PET/CT for diagnosis and treatment monitoring of inflammatory and infectious diseases. Clin Dev Immunol. 2013; 623036. doi: 10.1155/2013/623036.
Gotthardt M, Bleeker-Rovers CP, Boerman OC, Oyen WJ. Imaging of inflammation by PET, conventional scintigraphy, and other imaging techniques. J Nucl Med. 2010; 51(12): 1937-1949. doi: 10.2967/jnumed.110.076232.
Granger DN, Senchenkova E (eds.): Inflammation and the Microcirculation. Morgan & Claypool, 2010. NBK53373.
Haas H, Petrik M, Decristoforo C. An iron-mimicking, Trojan horse-entering fungi - has the time come for molecular imaging of fungal infections? PLoS Pathol. 2015; 11(1): e1004568. doi: 10.1371/journal.ppat.1004568.
Hammoud DA. Molecular imaging of inflammation: current status. J Nucl Med. 2016; 57: 1161-1165. doi: 10.2967/jnumed.115.161182
Israel O, Keidar Z. PET/CT imaging in infectious conditions. Ann N Y Acad Sci. 2011; 1228: 150-166. doi: 10.1111/j.1749-6632.2011.06026.x.
Kumar V, Boddeti DK. 68Ga-radiopharmaceuticals for PET imaging of infection and inflammation. Recent Results Cancer Res. 2013; 194: 189-219. doi: 10.1007/978-3-642-27994-2_11.
Lahoz-Beneytez J, Schnizler K, Eissing T. A pharma perspective on the systems medicine and pharmacology of inflammation. Math Biosci. 2015; 260: 2-5. doi: 10.1016/j.mbs.2014.07.006.
Nielsen OL, Afzelius P, Bender D, Schønheyder HC, Leifsson PS, Nielsen KM, Larsen JO, Jensen SB, Alstrup AK. Comparison of autologous 111In-leukocytes, 18F-FDG, 11C-methionine, 11C-PK11195 and 68Ga-citrate for diagnostic nuclear imaging in a juvenile porcine haematogenous staphylococcus aureus osteomyelitis model. Am J Nucl Med Mol Imaging 2015; 5(2): 169-182. PMCID: PMC4396013.
van Oosten M, Hahn M, Crane LMA, Pleijhus RG, Francis KP, van Dijl JM, van Dam GM. Targeted imaging of bacterial infections: advances, hurdles and hopes. FEMS Microbiol Rev. 2015; 39(6): 892-916. doi: 10.1093/femsre/fuv029.
Petrik M, Zhai C, Haas H, Decristoforo C. Siderophores for molecular imaging applications. Clin Transl Imaging 2017; 5(1): 15-27. doi: 10.1007/s40336-016-0211-x.
Riccardi C, Levi-Schaffer F, Tiligada E (eds.): Immunopharmacology and Inflammation. Springer, 2018. ISBN 978-3-319-77658-3. doi: 10.1007/978-3-319-77658-3.
Roivainen A, Jalkanen S, Nanni C. Gallium-labelled peptides for imaging of inflammation. Eur J Nucl Med Mol Imaging 2012; 39(Suppl 1): S68-S77. doi: 10.1007/s00259-011-1987-6.
Roivainen A, Saraste A. In vivo imaging of inflammation. In: Cavaillon J-M, Singer M (eds.) Inflammation: from molecular and cellular mechanisms to clinic. Wiley, 2017, pp 1567-1582. doi: 10.1002/9783527692156.ch62.
Rowe SP, Cho SY. The role of PET in the evaluation of musculoskeletal infections. Semin Musculoskelet Radiol. 2014; 18: 166-174. doi: 10.1055/s-0034-1371018.
Signore A, Glaudemans AW. The molecular imaging approach to image infections and inflammation by nuclear medicine techniques. Ann Nucl Med. 2011; 25(10): 681-700. doi: 10.1007/s12149-011-0521-z.
Signore A, Glaudemans AW, Galli F, Rouzet F. Imaging infection and inflammation. Biomed Res Int. 2015; 615150. doi: 10.1155/2015/615150.
Sugawara Y, Gutowski TD, Fisher SJ, Brown RS, Wahl RL. Uptake of positron emission tomography tracers in experimental bacterial infections: a comparative biodistribution study of radiolabeled FDG, thymidine, L-methionine, 67Ga-citrate, and 125I-HSA. Eur J Nucl Med. 1999; 26(4): 333-341. doi: 10.1007/s002590050395.
Wagner T, Basu S (eds.): PET/CT in Infection and Inflammation. Springer, 2018. doi: 10.1007/978-3-319-90412-2.
Wu C, Li F, Niu G, Chen X. PET imaging of inflammation biomarkers. Theranostics 2013; 3(7): 448-466. doi: 10.7150/thno.6592.
Updated at: 2022-03-19
Created at: 2015-08-18
Written by: Vesa Oikonen