Analysis of [18F]FDOPA PET data

L-Dihydroxyphenylalanine (L-DOPA) is a precursor of catecholamines (dopamine, noradrenalin, and adrenalin). It crosses the blood-brain barrier (BBB) by facilitated diffusion, mainly by L-type amino-acid transporter (LAT1), and its conversion to dopamine is catalysed by aromatic amino acid decarboxylase (AADC). L-DOPA in the brain is mainly synthesized by hydroxylation of tyrosine. Dopamine is stored in the vesicles of the dopamine neurons. There are several L-DOPA analogues available for PET imaging, including 6-[18F]-L-DOPA (6-[18F]fluoro-l-3,4-dihydroxy-phenylalanine, [18F]FDOPA), 2-[18F]-L-DOPA (2-[18F]fluoro-l-3,4-dihydroxy-phenylalanine), [18F]FBPA (4-borono-2-18F-fluoro-l-phenylalanine), and [18F]FPDOPA (N-(2-[18F]fluoropropionyl)-3,4-dihydroxy-L-phenylalanine).

L-DOPA and 6-[18F]FDOPA

6-[18F]-L-DOPA (FDOPA) can be used to study the dopaminergic function in the brain, since its uptake reflects the activity of AADC and subsequent build-up of labelled metabolites fluorodopamine (FDA), 6-[18F]fluorohomovanillic acid (FHVA), and 6-[18F]fluoro-3,4-dihydroxyphenylacetic acid (FDOPAC) (Firnau et al., 1987; Barrio et al., 1997). FDOPA is also the most used PET tracer for studying hyperinsulinemia and neuroendocrine tumours (NETs).

[18F]FDOPA in tumour imaging

Tumour cell growth can be assessed using labelled amino acids, and [18F]FDOPA as an amino-acid analogue can be used for that purpose. Tumours generally have upregulated amino-acid transport, not only because of increased demand for protein synthesis, but also for usage as substrate for energy production. Incorporation of labelled amino-acid in other metabolic pathways than protein synthesis is not a drawback.

Transport of amino-acids into cells is not a rate-limiting step in peripheral tissues and tumours, but in the brain the uptake may be limited by the capacity of the carrier system or partial saturation by physiological amino-acids in plasma (Vaalburg et al., 1992). High molar activity may slightly increase SUV of NETs (Stormezand et al., 2021).

Blocking of [18F]FDOPA metabolism may affect the tumour-to-background ratio, not necessarily for the better (Tuomela et al., 2013).

The other L-DOPA analogues than the 6-[18F]-L-DOPA have been mainly used for tumour imaging. 2-[18F]-L-DOPA has higher uptake in melanoma than 6-[18F]-L-DOPA (Ishiwata et al., 1989 and 1991). [18F]FBPA was developed for BNCT but has been used in diagnostics, too.

Brain tumours

LAT1 is upregulated in the tumour tissue and endothelium, and the lower pH at the site of tumour also increases the LAT1 activity. Labelled amino-acid analogues are particularly useful for imaging brain tumours, because amino-acid uptake in the normal brain tissue is relatively low (Papin-Michault et al., 2016). [11C]methionine and [18F]fluoroethyl-L-tyrosine are commonly used tracers that also are transported by LAT1. Although correlation between LAT1 expression and SUV of amino-acid tracers has been seen in some studies, only minimal LAT1 expression is required for high [18F]FDOPA uptake in tumours. In addition, LAT4 may transport [18F]FDOPA. Different uptake mechanisms can occur among different tumours even within an individual (Feral et al., 2017). 3-O-methyl-FDOPA, one of the main metabolites of FDOPA, is also a substrate of LATs.

In assessment of gliomas by dynamic [18F]FDOPA PET, combining simple parameters from the uptake dynamics can predict isocitrate dehydrogenase (IDH) mutations and 1p/19q codeletion (Ginet et al., 2020; Zaragori et al., 2020 and 2022). IDH-mutant gliomas have higher LAT1 expression and [18F]FDOPA uptake (Cobes et al., 2023).

Functional brain PET

[18F]FDOPA uptake could be used as an index of dynamic regulation of dopamine synthesis enzymes by neuronal firing. Bolus+infusion protocol has been used to assess task-specific changes in dopamine synthesis (Hahn et al., 2021).

Drawbacks of FDOPA

Peripheral metabolism of FDOPA leads to considerable concentrations of 18F-carrying metabolites in the plasma. The main metabolites in plasma are FDA, 3-O-sulphato-FDOPA, and 3-O-methyl-FDOPA (OMFD) (Firnau et al., 1988; Tuomela et al., 2013). OMFD can pass the blood-brain barrier (BBB) via the same amino acid transporters as FDOPA. The rate of peripheral metabolism and relative proportions of plasma metabolites is affected by enzyme inhibitors (such as carbidopa) that are commonly administered before FDOPA PET study, or that are routinely used with L-DOPA medication (Tuomela et al., 2013). Carbidopa pretreatment increases plasma [18F]FDOPA concentration, and consequently the radioactivity concentration in the brain (Melega et al., 1990), but does not affect the Patlak plot influx rate (Hoffman et al., 1992). In glioma studies, the carbidopa pretreatment increases SUV in tumour and healthy brain, and does not affect the tumour-to-healthy-brain ratio (Bros et al., 2021).

Transport of FDOPA and OMFD across BBB is affected by the competition with amino acids in the plasma and brain. DOPA is synthesized endogenously in the brain. Under normal physiological conditions the FDOPA transport is close to saturation for both influx and efflux, and K1 and k2 are approximately inversely proportional to the concentration of large neutral amino acids in the plasma and brain, respectively (Cunningham and Lammertsma, 1989). Leenders et al (1986) have shown that intravenous amino acid loading reduced FDOPA brain uptake about 3-fold compared to the fasting conditions.

Transport of amino acids from plasma to red blood cells is slow. For L-DOPA the equilibration half-life is ∼1 h (Floud & Fahn, 1981).

[18F]FDOPA and its metabolites are excreted into urine, and accumulation is very high in the kidneys and urinary bladder (Ribeiro et al. 2005), which may be a problem in studying these organs, prostate, and the tail of the pancreas.

Analysis methods used in literature

Gjedde-Patlak graphical analysis with arterial plasma input

Multiple-time graphical analysis for irreversible kinetics (Patlak plot) has frequently been used to assess the net influx rate constant Ki, representing dopamine synthesis and storage in the brain. While the synthesized [18F]dopamine is stored in the presynaptic vesicles, some of it is metabolized further into [18F]FHVA and [18F]FDOPAC, which can leave the brain, causing underestimation of Ki without observable downward curvature, even when only 90 min of the data is used (Sossi et al., 2003). Brain scans are typically 90 min long, and the time range for Patlak plot line fit is often set to 30-90 min. However, the applied time range for line fit has been varied in the numerous FDOPA studies applying the Patlak plot, for example, time range 20-70 min was used by Heinz et al (2005).

Peripherally formed labelled metabolites of [18F]FDOPA, especially [18F]OMFD, can penetrate BBB, which must be accounted for in the analysis. The radioactivity concentration in a reference region with little or no AADC activity (cerebellum or occipital cortex) consists mainly of plasma-derived [18F]FDOPA and [18F]OMFD. Reference region TAC is subtracted from the TAC of region-of-interest (ROI) or TACs of image pixels; after the subtraction the remaining radioactivity concentration in region of interest represents the concentration of [18F]FDA and its metabolites, enabling better assessment of AADC activity (Martin et al., 1989). Plasma TAC must be corrected for all labelled metabolites.

In regional analysis the Patlak net influx rate constant Ki can be calculated per striatum, in units (mL plasma)×(striatum×min)-1. In this case, striatal ROI is drawn so that it covers the the whole striatum as visible in the PET image, including any spill-out activity; the regional radioactivity concentrations (C) and the volumes (V) or areas of the striatal and reference tissue regions are recorded, and the "specific" striatal radioactivity (A) is then calculated at each time point as

(Martin et al., 1989). This approach may help to reduce the partial volume effect, and may be clinically more relevant measure of nigrostriatal function than the striatal average; the caudate, putaminal, and thalamic volumes are negatively correlated with age, and reduced in Parkinson's disease (Lisanby et al., 1993). Usually, though, the subtraction is done simply using regional average TACs, providing also the Patlak net influx rate constant Ki in usual units (mL plasma)×(ml tissue)-1×min-1. Then the ROIs are usually drawn based on anatomical reference image.

Graphical analysis with reference tissue input

Cerebellum and occipital cortex are used as reference regions in brain [18F]FDOPA studies. Multiple-time graphical analysis for irreversible kinetics (Patlak plot) can be used with either plasma input function or with reference tissue input function (reference tissue MTGA). Patlak plot with reference tissue input has been used for example to assess the rate of loss of nigrostriatal dopaminergic neurons in Parkinson's disease has been measured using Patlak plot with occipital cortex as reference region (Nurmi et al., 2001; Brück et al., 2009).

In case of the reference tissue input, the slope of the Patlak plot Kiref is a function of k2 and k3:

The effect of competing amino acids in the brain is cancelled out from Kiref results, because the competition affects similarly the reference region and the region of interest (ROI).

The loss of [18F]dopamine metabolites from the striatum causes underestimation of Patlak plot slope, even more so with reference tissue input than with plasma input (Sossi et al., 2003). Presence of [18F]OMFD introduces downward curvature in the Patlak plot, causing negative bias in Kiref. The negative bias in amplified with the loss of dopaminergic neurons, which may explain why reference region input Patlak has been often found to be the most sensitive marker for Parkinson's disease (Sossi et al., 2003).

Dual time point imaging can be used to calculate a surrogate parameter to Kiref (Alves et al., 2017).

Standardized uptake value (SUV)

The semiquantitative SUV method is frequently used in oncological applications where absolute quantification is not necessary. Ribeiro et al. (2005) showed that the radioactivity concentration in pancreas, liver and lung remained rather constant (although small decrease can be observed) between 50 and 80 min after administration of [18F]FDOPA. This suggests that dynamic PET scanning is not necessary, but a static 5-min scan between 45 and 90 min should be informative enough (Becherer et al. 2004; Ribeiro et al. 2005). Although the concentration decrease with increasing time was small, it suggests that reversible uptake models (e.g. Logan graphical analysis) measuring distribution volume would be more appropriate for the analysis than irreversible uptake models (Patlak plot). If arterial plasma curve is not measured, a tissue-to-reference tissue (preferably tissues with high perfusion and blood content but low AADC activity) ratio approach may be related to distribution volume and be proven useful.

Tumour-to-normal tissue ratio (T/N) can be calculated from SUVs or directly from the regional radioactivity concentrations, and is also often used, especially in analysis of the brain, where the healthy hemisphere can be used as reference tissue.

Tissue-to-blood ratio of FDOPA has been used to assess sympathetic activity in myocardium (Burger et al., 2018).

Simplified kinetic analysis

For assessing gliomas, a 20-min dynamic PET scan was found to provide parameters (K1, k2, and time-to-peak) that could separate high- and low-grade gliomas, unlike parameters from static 10-30 min scan; a reversible one-tissue compartmental model with vascular volume fraction as third parameter was applied to the data, using image-derived input function with literature-based metabolite correction (Girard et al., 2021).

See also:


Barrio JR, Huang SC, Phelps ME. Biological imaging and the molecular basis of dopaminergic diseases. Biochem Pharmacol. 1997; 54: 341-348. doi: 10.1016/S0006-2952(97)00031-2.

Cumming P: Imaging Dopamine. Cambridge University Press, 2009. ISBN: 9780521790024. doi: 10.1017/CBO9780511575853.

Cunningham VJ, Lammertsma AA. Kinetic modelling of L-18F-fluorodopa uptake for PET studies in man. In: Positron Emission Tomography in Clinical Research and Clinical Diagnosis: Tracer Modelling and Radioreceptors, edited by C Beckers et al., 1989, 81-89. ISBN 978-0-7923-0254-4.

Firnau G, Garnett ES, Chirakal R, Sood S, Nahmias C, Schrobilgen G. [18F]Fluoro-L-dopa for the in vivo study of intracerebral dopamine. Int J Rad Appl Instrum A 1986; 37(8): 669-675. doi: 10.1016/0883-2889(86)90260-1.

Firnau G, Sood S, Chirakal R, Nahmias C, Garnett ES. Cerebral metabolism of 6-[18F]fluoro-L-3,4-dihydroxyphenylalanine in the primate. J Neurochem. 1987; 8(4): 1077-1082. doi: 10.1111/j.1471-4159.1987.tb05629.x.

Firnau G, Sood S, Chirakal R, Nahmias C, Garnett ES. Metabolites of 6-[18F]fluoro-L-dopa in human blood. J Nucl Med. 1988; 29(3): 363-369. PMID: 3126278.

Heinz A, Siessmeier T, Wrase J, Buchholz HG, Gründer G, Kumakura Y, Cumming P, Schreckenberger M, Smolka MN, Rösch F, Mann K, Bartenstein P. Correlation of alcohol craving with striatal dopamine synthesis capacity and D2/3 receptor availability: a combined [18F]DOPA and [18F]DMFP PET study in detoxified alcoholic patients. Am J Psychiatry 2005; 162(8): 1515-1520. doi: 10.1176/appi.ajp.162.8.1515.

Huang SC, Yu DC, Barrio JR, Grafton S, Melega WP, Hoffman JM, Satyamurthy N, Mazziotta JC, Phelps ME. Kinetics and modeling of L-6-[18F]fluoro-dopa in human positron emission tomographic studies. J Cereb Blood Flow Metab. 1991; 11(6): 898-913. doi: 10.1038/jcbfm.1991.155.

Kienast T, Hariri AR, Schlagenhauf F, Wrase J, Sterzer P, Buchholz HG, Smolka MN, Grunder G, Cumming P, Kumakura Y, Bartenstein P, Dolan RJ, Heinz A. Dopamine in amygdala gates limbic processing of aversive stimuli in humans. Nat Neurosci. 2008; 11(12): 1381-1382. doi: 10.1038/nn.2222.

Kumakura Y, Cumming P, Vernaleken I, Buchholz H-G, Siessmeier T, Heinz A, Kienast T, Bartenstein P, Gründer G. Elevated [18F]Fluorodopamine turnover in brain of patients with Schizophrenia: an [18F]Fluorodopa positron emission tomography study. J Neurosci. 2007; 27(30): 8080-8087. doi: 10.1523/JNEUROSCI.0805-07.2007.

Kumakura Y, Danielsen EH, Gjedde A, Vernaleken I, Buchholz HG, Heinz A, Gründer G, Bartenstein P, Cumming P. Elevated [18F]FDOPA utilization in the periaqueductal gray and medial nucleus accumbens of patients with early Parkinson's disease. Neuroimage 2010; 49(4): 2933-2939. doi: 10.1016/j.neuroimage.2009.11.035.

Leenders KL, Poewe WH, Palmer AJ, Brenton DP, Frackowiak RSJ. Inhibition of L-[18F]fluorodopa uptake into human brain by amino acids demonstrated by positron emission tomography. Ann Neurol. 1986; 20: 258-262. doi: 10.1002/ana.410200212.

Martin WR, Palmer MR, Patlak CS, Calne DB. Nigrostriatal function in humans studied with positron emission tomography. Ann Neurol. 1989; 26(4): 535-542. doi: 10.1002/ana.410260407.

Misu Y, Goshima Y (eds.): Neurobiology of DOPA as a Neurotransmitter. CRC Press, 2006. ISBN: 978-0-415-33291-0.

Tuomela J, Forsback S, Haavisto L, Vahlberg T, Grönroos TJ, Solin O, Haaparanta-Solin M. Enzyme inhibition of dopamine metabolism alters 6-[18F]FDOPA uptake in orthotopic pancreatic adenocarcinoma. EJNMMI Res. 2013; 3: 18. doi: 10.1186/2191-219X-3-18.

Vaalburg W, Coenen HH, Crouzel C, Elsinga PH, Långström B, Lemaire C, Meyer GJ. Amino acids for the measurement of protein synthesis in vivo by PET. Int J Rad Appl Instrum B. 1992; 19(2): 227-237. doi: 10.1016/0883-2897(92)90011-M.

Tags: , , ,

Updated at: 2023-11-17
Created at: 2009-02-18
Written by: Vesa Oikonen