Abstract
Objective: To evaluate the visible and quantitative anatomic distribution of fluorine-18–labeled l-DOPA in the healthy human brain, to thereby expand the understanding of extrastriatal sites of levodopa function, and to provide a broader foundation for clinical and research studies of fluoroDOPA accumulation in patients.
Methods: The authors performed dynamic three-dimensional fluoroDOPA PET imaging in 10 healthy volunteers and analyzed the images visually and quantitatively. Twenty-eight regions of interest were applied to parametric images of the uptake rate constant (using the multiple-time graphic plot method with cortical input function) and also were used to quantitate regional radioactivity at 80 to 90 minutes. The authors correlated the uptake constants with published human regional neurotransmitter and decarboxylation data.
Results: PET imaging with fluoroDOPA demonstrates trapping of labeled dopamine or its metabolites in substantial quantities in many areas of the brain other than the mesostriatal pathways, including considerable uptake in the serotonergic and noradrenergic areas of the hypothalamus and brainstem as well as in extrastriatal cerebral sites. Total fluoroDOPA uptake correlates best with the sum of catecholamine and indolamine concentrations in the brain and moderately well with regional activity of aromatic l-amino acid decarboxylase, but correlates poorly with extrastriatal dopamine concentration.
Conclusion: Neither l-DOPA nor its radiolabeled analog fluoroDOPA is metabolized or accumulates specifically in dopaminergic or even catecholaminergic neurons. Substantial dopamine production within serotonin and norepinephrine neurons may play a role in either therapeutic effects or adverse effects of therapy with l-DOPA.
After years of its routine use in the treatment of PD, levodopa sometimes is viewed only as a means of replacing the dopamine lost when nigrostriatal neurons degenerate. However, the sites of conversion to dopamine, which play a role in the therapeutic—and adverse—effects of levodopa, are not fully known,1 limiting our understanding of levodopa use and the therapeutic differences between levodopa and direct dopamine receptor agonists. Whereas animal studies show that l-DOPA and its analogs are decarboxylated in nondopaminergic cells and in extrastriatal brain regions, the diagnostic and scientific benefits of measuring this extrastriatal dopamine in living humans have not been extensively explored. In this study, we use PET to evaluate the sites of exogenous DOPA localization in the healthy human brain.
Firnau et al. proposed the use of fluorine-18–labeled l-DOPA as an in vivo marker of dopamine metabolism over 25 years ago.2 For more than 15 years,3 6-fluoroDOPA PET has been used to provide indices of exogenous l-DOPA decarboxylation and storage in humans. It appears in approximately 300 published studies of the mesostriatal dopaminergic system and may have a clinical role in the diagnosis and differential diagnosis of parkinsonism. Although some of these publications mention extrastriatal specific uptake, there is no systematic evaluation of the extrastriatal accumulation of fluoroDOPA activity in the PET literature. Many of these reports also refer to fluoroDOPA as a label for aromatic l-amino acid decarboxylase (AAAD), but there is little independent evidence on which to judge the similarity of localization of fluoroDOPA and AAAD activity. AAAD and the catecholamine and indolamine neural systems are present throughout the brain and are thought to play roles in psychiatric diseases,4 limbic functions,5 and general modulation of cerebral activity as well as parkinsonism; therefore, imaging of their extrastriatal components is potentially as important as striatal imaging. In this study, we evaluate the measuring of AAAD and amine neurotransmitters with fluoroDOPA by correlating PET findings with published human in vitro data.
Subjects and methods.
Subjects.
The 10 human subjects in this report (6 women, 4 men; mean age 62 years, range 49 to 79 years) are healthy control subjects enrolled in a longitudinal study of the functional chemoanatomy of PD. None has developed neurologic disease during the 3 to 5 years in this study. These studies were approved by the University of Wisconsin Health Sciences Human Subjects Committee and Radioactive Drug Research Committee. We obtained informed consent from all subjects.
PET imaging.
We prepared fluoroDOPA according to the method of Namavari et al.6 using stannylated precursor material purchased from Boz Chemical Engineering (Montreal, Quebec, Canada). The radiochemical purity of fluoroDOPA was greater than 95%, and the specific activity was 400 to 800 mCi/mmol. We performed PET imaging on an Advance scanner (General Electric Medical Systems, Waukesha, WI) whose characteristics have been described.7 The intrinsic resolution in the central 10 cm of this instrument is 4 to 5 mm. We acquired all scans in three-dimensional mode.
Subjects had a low-protein breakfast 4 to 6 hours before the scan to minimize competition from dietary amino acids. They received carbidopa (100 mg orally) 1 hour before radiopharmaceutical injection to block peripheral decarboxylation of fluoroDOPA. Subjects were positioned on the PET scanning couch in a comfortable supine position, with the head lightly restrained in a head holder to minimize motion during imaging. We obtained a 20-minute transmission scan using rotating 68Ge pin sources to correct for photon attenuation, then began a 90-minute dynamic PET acquisition simultaneous with IV administration of 204 to 284 MBq (5.5 to 7.7 mCi) of fluoroDOPA. This scan consisted of 18 time frames with durations lengthening progressively from 30 seconds initially to 10 minutes at the end. We reconstructed the data using a 24-cm field of view, Hanning transaxial filter (4-mm cutoff), ramp axial filter (8.5-mm cutoff), and 128 × 128 image matrix.
Analysis.
For qualitative analysis, we summed the images acquired 30 to 90 minutes after injection into a single-image set for each subject. We chose this time period to include contributions from both earlier data (when greater intravascular activity provides better anatomic detail for the whole brain) and later data (when specific accumulation in monoamine-avid structures predominates). Brain structures large enough to be resolved by PET imaging were identifiable directly from these images, as were structures of any size that had adequate specific uptake. We generated tracer uptake rate constants (Kc) for quantitative analyses using occipital (calcarine) cortex activity as a tissue input function, using the graphical method described by Patlak and Blasberg.8 We generated Kc parametric images by performing the graphical analysis for each image voxel.
We selected anatomic regions of interest based on imaging evidence of specific fluoroDOPA uptake or published data indicating increased regional decarboxylase activity or monoamine neurotransmitter concentration. The cerebellum, with its low monoaminergic innervation, was included because it is commonly used as a reference region in PET studies. Regions of interest were drawn on the summed images by a neuroradiologist experienced in brain imaging with PET (W.D.B.) and then applied to quantitative late-activity (80 to 90 minutes) and Kc images. We normalized each late-activity value by the sum of region of interest values for that subject to adjust for administered dose and global differences, and then expressed the regional means and standard deviations in terms of a fraction of putaminal activity.
We combined amine neurotransmitter concentration data from numerous sources9-21 because of limitations of the individual published reports. Each report we identified includes only a few selected brain regions, data from only a few cadavers, or assays of only one or two of the relevant neurotransmitters. The problems caused by early postmortem catecholamine and enzyme instability22 compounded the usual difficulties inherent in human autopsy studies. The various laboratories used different measurement methods. Despite these difficulties, we found the relative concentrations of a given transmitter among various brain regions to be remarkably consistent among these reports. Because none of the absolute concentration values is apparently more correct than another, we chose to take the average of all available measurements as the most reliable available data set. The published data regarding AAAD activity in the human brain come from only one laboratory; we combined complementary data from two reports21,23 for our analysis. These studies showed large global intersubject variation but generally consistent patterns among brain regions,21 allowing the relative regional values to be combined. The tabulation of these data is on file with the National Auxiliary Publications Service (see Note at end of article).
Results.
Qualitative assessment of fluoroDOPA distribution.
Figures 1 through 3⇓⇓ demonstrate the distribution of activity in the summed images. A highly structured pattern of uptake is visible in the posterior fossa. The lower medulla appears homogeneous and similar in activity to cerebellar gray matter (figure 1A). In the upper medulla, there is greater tracer accumulation, restricted to the more central and posterior medullary areas; that is, it excludes the pyramidal tracts and inferior cerebellar peduncles. There appears to be slightly increased activity in the anterolateral medulla corresponding roughly to the region of the olive, but limited resolution makes it impossible to localize this activity in the inferior olivary nucleus with certainty. Within the pons, there is marked heterogeneity of activity. Specific tracer accumulation comparable with that in the upper medulla is present throughout the pontine tegmentum (figures 1B and 2A⇓). There is much less activity in the basis pontis and no visible internal structure. Within the cerebellum, the hemispheric cortex, central white matter, and vermian cortex are distinct in a pattern reflecting the relative amounts of gray matter (with its greater blood volume) and white matter.
Figure 1. Transaxial summed fluoroDOPA PET images (30–90 min after injection). (A) Lower medulla and cerebellum; little specific activity is visible at this level. (B) Pons; the marked difference between basis pontis and pontine tegmentum is visible, as well as slight inferior temporal uptake. (C) Pontomedullary junction and region of the dorsal raphé nucleus and locus ceruleus; this image also shows amygdala, anterior cingulate, and substantia innominata uptake. (D) Lower midbrain, rostral hypothalamus, and ventral striatum specific uptake. (E) Specific uptake in the upper midbrain, hypothalamus, septal area, globus pallidus, and inferior neostriatum. (F) Heterogeneous thalamic uptake and the pineal body are visible in addition to the caudate heads and putamina.
Figure 2. Sagittal reformatted fluoroDOPA images at midline (A), paramedian (B), and striatal (C) levels. The pattern of brainstem activity is well demonstrated, as well as that of the hypothalamus, cingulate gyrus, gyrus rectus, and subcallosal area.
Figure 3. Coronal reformatted images. (A) Through the genu of the corpus callosum; relatively high subcallosal and cingulate uptake are well seen. (B) At the level of the interventricular foramina of Monro; heterogeneous thalamic activity is visible, as well as prominent hypothalamic and anterior striatal uptake. Less prominent specific uptake is also visible in the amygdalae and in the cingulate gyri.
There is a marked focal increase in specific uptake at the junction of the pontine tegmentum and the dorsal midbrain at the expected location of the dorsal raphé nucleus and nearby locus ceruleus (figures 1C and 2A⇑). High activity throughout the dorsal midbrain appears to encompass the entire tectum, the periaqueductal gray matter, and part of the reticular formation (figures 1D and 2B⇑). In the center of the tegmentum, there is a smaller degree of specific uptake. This uptake is likely within the neurons of the ventral tegmental area; this region also contains the red nuclei and the decussation of the superior cerebellar peduncles, but individual structures are not resolved. On either side of the interpeduncular fossa, specific uptake in the substantia nigra appears approximately equal to that in the dorsal midbrain.
Separation between the substantia nigra and the substantial specific uptake of the hypothalamus is indistinct (figure 1, D and E; see figure 2A). The anterior hypothalamus has activity approximately equal to the greatest midbrain activity; posterior hypothalamic uptake is slightly lower. The thalamus shows relatively high specific uptake medially, posteriorly, and posterosuperiorly, in distinct contrast to the lower activity in its anterolateral and central portions (figure 1F). This pattern corresponds with the relative amounts of gray matter and myelinated fibers within these thalamic regions as seen in myelin stains, CT, and MR scans; therefore, its significance regarding specific tracer uptake in individual thalamic nuclei is uncertain. Medial thalamic activity appears continuous with that in the posterior hypothalamus and tectal plate because of their proximity. The activity relationships among the thalamic regions, hypothalamus, and dorsal midbrain are well appreciated on reformatted coronal images (figure 3B). Specific uptake also is visible in the pineal body (figure 1F) and in the pituitary infundibulum.
The activity of the anterior hypothalamus appears to blend smoothly through the septal nucleus area into the nucleus accumbens septi anteriorly and the globus pallidus laterally (figure 1, D and E). The globus pallidus is distinctly visible as an area of relatively low specific uptake between the very low activity of the internal capsule and the very high activity of the putamen. The region, which includes the nucleus accumbens and the ventral striatum/substantia innominata has specific uptake that is greater than that of the diencephalic structures but less than that of the neostriatum (figure 1, C and D.) The highest uptake of fluoroDOPA is in the dopamine-rich caudate nucleus and putamen; tracer accumulation appears slightly higher in the putamen (figure 1F). Specific uptake is visible in the caudate body to approximately the coronal level of the pulvinar (figure 2C).
Among extrastriatal cerebral structures, there is substantial specific uptake in the amygdala and in the hippocampal formation (see figure 1C). The amygdaloid body has slightly lower activity than the midbrain; it is easily identified by its shape and relationship to the temporal lobe uncus. There is apparently slightly less activity within the hippocampal formation. This structure is best seen at its larger head region but in many scans can be followed through the hippocampal body to the coronal level of the posterior midbrain. The parahippocampal gyrus and its extension into the isthmus of the cingulate gyrus show slightly greater uptake than the surrounding neocortex.
Of the remainder of the cerebral cortex, the subcallosal area and gyrus rectus appear to have the greatest specific tracer uptake, followed by the anterior cingulate gyrus. These are best appreciated in sagittal and coronal reformatted images (figures 2B and 3A⇑). The medial surface of the frontal lobe outside of the cingulate sulcus shows slightly higher activity than more lateral frontal neocortex. The prefrontal areas as a whole have slightly greater activity than the posterior frontal, parietal, and lateral temporal cortices. Williams and Goldman-Rakic describe a similar pattern of frontal lobe dopamine concentration in monkeys.24 The occipital pole and posterolateral occipital cortex appear to have slightly less activity than other neocortical areas. The summed images show slightly higher activity in the calcarine cortex than in the lateral occipital lobe, apparently because of the large intravascular volume of this tissue and the inclusion of early data in the summed images.
Quantitative evaluations.
We tabulated two regional quantitative PET analyses: normalized radioactivity at 80 to 90 minutes after injection, and regional uptake constants (Kc). This table is filed with National Auxiliary Publications Service (see Note at end of article). Standard deviations were included to indicate measurement reliability among subjects. Whereas simple regional radioactivity includes the homogeneous background caused by 3-O-methyl-fluoroDOPA, it is useful for analyses based on ratios and provides normal data with which patient studies may be compared. Kc reflects specific uptake and long-term storage.
Some aspects of the pattern visible in these quantitative data are well known, others less so. The highest uptake rate constant was in the putamen. The other basal nuclear structures known to be rich in dopamine also showed high uptake, ranging from 40% of putaminal uptake in the septal nuclei area and 75% in the nucleus accumbens, to 94% in the caudate head. In the brainstem the uptake constants in the medulla and the pontine tegmentum were approximately 20% of putaminal uptake, whereas all midbrain regions were in the range of 35% to 40%. The caudal hypothalamus was slightly lower (25%). The medial and posterior portions of the thalamus had uptake constants similar to those in the lower brainstem; the rostral hypothalamus and globus pallidus showed specific uptake similar to that in the midbrain. The uptake rate constant for the amygdala also was 35% of the putaminal constant, and the hippocampal head and body, slightly lower at almost 30%. Frontal and temporal cortices were in the range of 10% to 15% of putaminal uptake, and the anterior cingulate, slightly higher at about 20%. Specific uptake in the parietal cortex was near zero. The cerebellum (both gray matter and white matter) had Kc values indistinguishable from zero, as did the lateral thalamus and the basis pontis. The centrum semiovale white matter artifactually showed an uptake value of about 15% that of the putamen because of blurring of adjacent caudate and putamen activity into this region.
Correlations between corresponding data from the described neurochemical and PET tables are presented graphically in figure 4. These plots show poor correlation between fluoroDOPA specific uptake and dopamine concentration overall; there is a wide range of Kc values for regions with low dopamine concentrations. FluoroDOPA specific uptake correlates best with the total concentration of these four AAAD-dependent monoamine neurotransmitters. The correlation of Kc with regional AAAD activity also is fair, which provides some support for the proposition that fluoroDOPA is largely a marker for AAAD activity. This correlation is likely underestimated because of the poor sensitivity and precision of the published AAAD assays. For example, these factors lead to clustering of points in a column at the limit of detectability of AAAD activity in figure 4C.
Figure 4. Correlations of fluoroDOPA regional uptake rate constants (Kc) with (A) dopamine concentrations; (B) the sum of local concentrations of dopamine, norepinephrine, epinephrine, and serotonin; and (C) AAAD activities. In A, the four points with dopamine concentration greater than 0.5 are, in order of increasing dopamine, substantia nigra, nucleus accumbens, caudate, and putamen. The other measured structures have a substantial range of Kc values but low dopamine content, giving this cluster a distinctly higher slope and indicating the large nondopaminergic contribution to fluoroDOPA uptake. B demonstrates that fluoroDOPA Kc correlates better with the summed concentration of all amine neurotransmitters than with that of dopamine alone. C shows a fair correlation of fluoroDOPA Kc with AAAD activity, but is limited by the quality of the autopsy enzyme measurements. The cluster of points with normalized AAAD activity of approximately 0.05 appears superficially similar to the low dopamine concentration cluster in A, but is an artifact of poor assay sensitivity.
Discussion.
6-FluoroDOPA has been extensively evaluated through the last two decades and consistently has been shown to follow the cellular and biochemical pathways of l-DOPA. These include transport across the blood–brain barrier and plasmalemma through the high-affinity large neutral amino acid carrier25; decarboxylation of the amino acid to form the amine26; vesicularization of the amine in the presynaptic terminal27; reuptake and exocytosis at the synapse28; and degradation of the amine by monoamine oxidase and catechol–O-methyltransferase (COMT) to form acid metabolites, which cross the blood–brain barrier into the plasma and undergo renal excretion.29 But the enzymes and transporters important to dopaminergic systems are more widely distributed than dopamine. AAAD is present in high concentrations in all catecholaminergic neurons and in even greater concentrations in serotonergic neurons.30 It also is found in the capillary endothelium and pericytes,31 in nonmonoaminergic neurons,1,30 and in glial cells.32 Decarboxylation of exogenous l-DOPA to dopamine has been demonstrated in serotonergic neurons.33
In 1984, Horne et al.34 published a study in rats in which they localized and chemically characterized the accumulation of l-DOPA products in catecholaminergic forebrain and midbrain regions using 14C-l-DOPA autoradiography, 3H-l-DOPA liquid radiochromatography, reserpine treatment, and selective lesioning with 6-hydroxydopamine. In addition to activity accumulation in dopamine-rich regions (striatum, nucleus accumbens, olfactory tubercle, and substantia nigra), the authors drew attention to specific radioactivity accumulation in structures in which norepinephrine predominates: the paraventricular and supraoptic hypothalamic nuclei and the medial geniculate body. Table 1 from that report also demonstrates varying amounts of specific 14C uptake in limbic structures, neocortical regions, and subsegments of the diencephalon and midbrain. In the current report, we have extended the previous work to include serotonin and AAAD chemically, and the brainstem anatomically, and have performed the studies in living humans. Because catecholamine concentrations far exceed that of serotonin in most of the midbrain and forebrain, our data are in general agreement with the earlier animal data.
Two limitations of using fluoroDOPA PET to study the fate of exogenous l-DOPA should be considered. First, exogenous l-DOPA leads to nonexocytotic release of dopamine and rapid metabolism of this neurotransmitter,35 apparently because vesicular transport or storage capacity is greatly exceeded by the ability of AAAD to act on the large amount of exogenous l-DOPA.36 In contrast, vesicularization is necessary to the long-term specific accumulation of activity in the brain after fluoroDOPA administration. Therefore, the long-term accumulation of DOPA metabolites may not perfectly reflect the sites of levodopa action. Second, in fluoroDOPA PET, as in the 14C-DOPA autoradiography work just described,34 the presence of large amounts of the 3-O–methylated metabolite produced by peripheral COMT limits the sensitivity of the method. This problem may be solved by the use of COMT inhibitors or a tracer such as 6-[18F]fluoro-m-tyrosine,37 which is not a substrate for COMT. This study also points out deficiencies in the published human data for regional monoamine concentrations and especially AAAD activity. The existing ex vivo data do not allow us to judge whether the need for vesicularization causes fluoroDOPA PET measurements to correlate better with amine neurotransmitter concentrations than with AAAD activity.
We have described and quantified specific fluoroDOPA uptake in areas of the brain that are known to have large concentrations of dopaminergic cell bodies or nerve terminals, and also in many brain regions with other monoamine neurotransmitters (particularly norepinephrine and serotonin) but no substantial dopaminergic innervation. A more detailed understanding of the sites of accumulation of radiolabeled l-DOPA in the normal and diseased human brain may increase our understanding of the many effects of therapeutic l-DOPA. This understanding of the heterogeneous but widespread accumulation of l-DOPA and its analogs also demonstrates the applicability of fluoroDOPA PET beyond Parkinson disease.
Acknowledgments
Supported by the National Institutes of Neurological Disorders and Stroke, grant R29 NS31612.
Acknowledgment
The authors gratefully acknowledge the invaluable assistance of PET Research Manger Joan M. Hanson.
Note. Readers can obtain 3 pages of supplementary material from the National Auxiliary Publications Service, 248 Hempstead Turnpike, West Hempstead, NY 11552. Request document no. 05518. This is not a multiarticle document. Remit with your order, not under separate cover, in US funds only, $15.00 for photocopies or $5.00 for microfiche. Outside the United States and Canada, add postage of $4.50 for the first 20 pages and $1.00 for each 10 pages of material thereafter, or $5.00 for the first microfiche and $1.00 for each fiche thereafter. There is a $25.00 invoicing charge on all orders filled before payment.
- Received October 6, 1998.
- Accepted March 4, 1999.
References
Letters: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.
You May Also be Interested in
Hemiplegic Migraine Associated With PRRT2 Variations A Clinical and Genetic Study
Dr. Robert Shapiro and Dr. Amynah Pradhan