Address correspondence and reprint requests to Dr. Miikka Korja, Department of Neurology, Turku University Hospital, Kiinamyllynkatu 4-8, P.O. Box 52, FI-20521 Turku, Finland; E-mail: email@example.com
Summary: Purpose: Unverricht-Lundborg disease (ULD) is currently classified as progressive myoclonus epilepsy. Myoclonus, the characteristic symptom in ULD, suggests that dopamine neurotransmission may be involved in the pathophysiology of ULD. Our purpose was to examine brain dopaminergic function in ULD patients.
Methods: Four genetically and clinically diagnosed ULD patients and eight healthy controls were scanned with [11C]raclopride-PET. PET images were coregistered to individual 1.5T MR images and region-of-interest analysis was performed for the striatum and thalamus. Standardized uptake values and individual voxel-wise binding potential maps of the patients and controls were also analyzed.
Results: ULD patients had markedly higher (31–54%) dopamine D2-like receptor availabilities than healthy controls in both the striatum and the thalamus. The proportionally highest binding potentials were detected in the thalamus. There were no significant differences in the cerebellar uptake of [11C]raclopride in ULD patients versus healthy controls. Voxel-based results were in accordance with the region-of-interest analysis.
Conclusions: These results suggest that dopaminergic modulation at the level of the striatum and thalamus could be a crucial factor contributing to the symptoms of ULD. In the light of our data, we propose that ULD with dopamine dysfunction and dyskinetic symptoms shares certain pathophysiological mechanisms with classical movement disorders. Future studies are therefore warranted to study the effect of dopaminergic pharmacotherapy in ULD.
Progressive myoclonic epilepsies (PMEs) are a heterogeneous group of neurological disorders, where the acronym PME is used for a combination of two clinically diverse and progressive neurological symptoms: myoclonic jerks and tonic–clonic seizures. Of these rare disorders, ULD (also known as EPM1 or Baltic PME) is the most common single cause of PME. In the late seventies, this autosomal recessively inherited disorder was recognized as a geographic cluster in Finland with a prevalence of one per 20,000 births (Norio and Koskiniemi, 1979). The age of symptom onset in ULD varies between 6 and 15 years (Koskiniemi, 1974). Symptoms of myoclonus, seizures, and progressive neurological dysfunction are progressive, although there is a remarkable variation in the severity and combination of symptoms among ULD patients (Genton et al., 1990).
Since the identification of the genetic locus for ULD (Lehesjoki et al., 1991), it has been possible to confirm the clinical diagnosis with an analysis of the mutations in the CSTB gene (Pennacchio et al., 1996), which encodes a cysteine protease inhibitor, cystatin B. The cystatin B protein binds to a light neurofilament (NF-L) protein (Di Giaimo et al., 2002), which is expressed, for example, in the dopaminergic neurons in the substantia nigra pars compacta (Mueller et al., 2004), and in the thalamus (Clinton et al., 2003) in humans. In addition, cystatin B binds to a β-spectrin protein (Di Giaimo et al., 2002), which together with α-spectrin associates to form biologically relevant tetramers and higher order oligomers (Byers and Branton, 1985; Liu et al., 1987). Interestingly, neurotoxin 6-hydroxydopamine (6OHDA) induced apoptosis in dopamine neurons is associated with the expression of a caspase cleavage product of α-spectrin (Oo et al., 2002). Caspase belongs to a cysteine protease family, which participates in apoptosis. Despite the described cystatin B interactions (Di Giaimo et al., 2002), there are no hypotheses explaining the role of the molecular interactions in the disease process.
Myoclonus of ULD patients consists of movement-activated and frequent stimulus-sensitive spontaneous myoclonic jerks. In the most common movement disorder, Parkinson's disease (PD), the dopaminergic nigrostriatal pathway is believed to be associated with the dysfunction of sequential movement processing. The extrapyramidal symptoms of PD can be effectively alleviated with the dopamine precursor, levodopa, which in turn may trigger abnormal involuntary movements, i.e., dyskinesias, including myoclonus (Klawans et al., 1975). Interestingly, a small study has indicated that levodopa treatment can have a dramatic positive effect on PME patients (presumably ULD patients), as two patients, who had been wheelchair-bound for years, were able to walk during the levodopa treatment (Leino, 1981). In addition, both movement-activated myoclonus and stimulus-sensitive myoclonus were totally blocked in a ULD patient with a direct D1- and D2-like dopamine receptor agonist, apomorphine, whereas patients with, e.g., Lafora disease, Kufs' disease, and juvenile neuroaxonal dystrophy, showed only the blockage of stimulus-sensitive myoclonus (Mervaala et al., 1990). These clinical studies suggest that dopaminomimetic drugs can improve the symptoms of ULD, and that there may be a deficit of dopaminergic neurotransmission in ULD.
In this paper, we describe the findings of positron emission tomography (PET) imaging of four ULD patients with radioactive D2-like dopamine receptor antagonist. On the basis of our results, we discuss about the possible role of dopamine in symptomatology of patients with ULD.
The study included four unrelated ULD patients (three women and one man) with a mean age of 32 years (women 21, 31, and 42 years; man 35 years), whose morphological, biochemical, and molecular pictures had led to the diagnosis of ULD. The patients represent all ULD and PME patients in the Hospital District of Southwest Finland, which is the second largest hospital district in Finland. The mean duration of the disease was 21 years (range: 10–31 years). The molecular diagnosis showed CSTB gene mutations in all patients (Laboratory of Medical Genetics, University of Helsinki). Generalized tonic–clonic seizures were reported as the first symptom in all patients, but they became rare or were completely controlled by pharmacological treatment. Myoclonic jerks as the first symptom were reported in none of the patients. Movement-activated and stimulus-sensitive spontaneous myoclonus was present in all patients in spite of the antiepileptic and symptomatic treatments, which comprised sodium valproate, clonazepam, levetiracetam, piracetam, and topiramate. None of the patients had had phenytoin as a regular antiepileptic medication. Three of the patients needed a wheelchair for moving longer distances, but they were able to walk short distances (< 25 m) slowly. The oldest patient could walk only a few steps with the help of another person, and thus needed constant help in daily activities. Other patients needed occasional assistance due to uncontrolled myoclonic jerks. In addition to myoclonic jerks, three of the patients had ataxia-like incoordination, intention-like tremor, and dysarthria. The patients were kept on their usual medication during the study. The age- and sex-matched control group consisted of eight healthy subjects (six women and two men) with a mean age of 28 years (range: 20–33 years). The Ethical Committee of Turku University Hospital approved the experimental protocol, and informed consent was obtained from every participating subject.
MR images were obtained with a Philips Gyroscan Intera 1.5 T CV Nova Dual scanner (Philips, Best, The Netherlands). MRI included axial T2-weighted (repetition time [TR] 4,896 ms; echo time [TE] 100 ms, slice thickness 5 mm) spin-echo images with a field of view (FOV) of 240 × 240 mm2, a matrix size of 288 × 368, and two excitations. Coronal FLAIR images (TR 11,000 ms; time inversion [TI] 2,800 ms; TE 140 ms; slice thickness 5 mm) with a FOV of 230 × 230 mm2, a matrix size of 206 × 256, two excitations, and axial T1-weighted 3D images (TR 25 ms; TE 5,5; slice thickness 1 mm) with a FOV of 256 × 256 mm2, and a matrix size of 400 × 400 were also obtained. An experienced neuroradiologist (R.P.) analyzed all the images for areas of signal change and atrophy on a four-point scale (no atrophy, mild, moderate, or severe).
Each subject underwent a PET scan with [11C]raclopride using a GE Advance PET Scanner (GE Medical Systems, Milwaukee, WI, U.S.A.) in 3D mode as described (Kaasinen et al., 2004). Briefly, after intravenous bolus injection of [11C]raclopride, the uptake of [11C]raclopride was measured for 59 min in 28 frames (15 × 1 min, 7 × 2 min, 6 × 5 min). The mean injected dose of [11C]raclopride was 204 ± 9 MBq. Dynamic [11C]raclopride images were realigned with individual T1 MRIs using SPM2 running on the Matlab 6.5 for Windows (Math Works, Natick, MA, U.S.A.). SPM2 realign parameters were estimated from the summated [11C]raclopride images (all frames). After realignment, regions-of-interest (ROIs) were delineated using Imadeus software (Forima Inc., Turku, Finland) on the MRIs, and time activity data were calculated using these ROIs. ROIs were delineated to the caudate nucleus and putamen (3–4 consecutive planes), the thalamus (3–4 consecutive planes on the striatal level), and the cerebellum (three consecutive planes). Regional BP values were calculated using the cerebellum as the reference region in the simplified reference tissue model (Lammertsma and Hume, 1996). The scan-rescan reproducibility for the measurements has been shown to be excellent for the striatum and good for the thalamus, with an absolute variability of BPs <8% (Hirvonen et al., 2003). For the reference region, standardized uptake values (SUVs) were calculated from cerebellar time-activity curves (kBq/ml) controlling for injected dose (MBq) and subject weight (kg) (mean value 30–59 min after injection).
To confirm the results of the ROI-based analysis, individual voxel-wise BP maps were created and compared between groups. Parametric images were created on the basis of linearized basis functions of the simplified reference tissue model (Gunn et al., 1997). Preprocessing of the BP images was performed using SPM2 (Friston et al., 1995) running on Matlab. Spatial normalization of the BP images was carried out using summated images (integral of all 28 frames) and a ligand-specific template. After spatial normalization, the BP images were smoothed with a 5-mm Gaussian kernel (voxel-size 2 × 2 × 2 mm). The voxel-based analysis was confined to the basal ganglia and thalamus, using explicit masking with a search volume of 15,468 voxels. Between-group comparisons of images in SPM were carried out with one-way ANOVA. T-contrasts were used, height threshold was p = 0.01, and extent threshold 50 voxels. Values are expressed as p values corrected for multiple comparisons.
The SPSS 11.0.2 for Mac Os X statistical software was used for all statistical analyses, except voxel-based analyses. Owing to the relatively small sample size, between-group comparisons were made with the nonparametric Mann–Whitney U two-independent-samples test with two-tailed exact significance. Within-group comparisons of left–right side differences of BPs were conducted using Wilcoxon's signed-rank test with two-tailed exact significance. Values are expressed as mean ± SD together with corresponding p values. The statistical significance level was set at p < 0.05.
Increased uptake of radioactivity in the ULD patients compared with the controls was seen in the striatum and thalamus (Table 1). When the striatum was further subdivided, both the putamen and the nucleus caudatus showed high uptake values in patients (Table 1). There were insignificant differences in cerebellar SUVs between patients and controls, as presented in Table 1, suggesting that the result in the striatum and thalamus was not due to differences in cerebellar binding of the tracer. The calculated mean BPs in the striatum and thalamus were 31–54% higher in the ULD group than in the control group (Table 1). The most significant relative increase in the [11C]raclopride binding was seen in the thalamus (54%) of the ULD group. Age did not correlate with [11C]raclopride binding in the striatum or thalamus, even though there was a slight trend toward declining BPs in every area with increasing age in both groups (Fig. 1). However, the oldest patient with the most severe myoclonus had the highest thalamic BP (Fig. 1). Hemispheric left–right side within-group comparison of BPs did not show significant (p > 0.05) differences.
Table 1. Mean values and spread for [11C]raclopride binding potentials in different brain regions. Relative increases of [11C]raclopride binding for each region are also presented
Nonparametric Mann–Whitney U two-independent-samples test with two-tailed exact significance was used for the statistical analysis. Cereb, cerebellum.
3.38 ± 0.16
0.60 ± 0.03
3.65 ± 0.11
2.93 ± 0.18
0.69 ± 0.10
2.50 ± 0.26
0.39 ± 0.05
2.79 ± 0.31
2.09 ± 0.27
0.68 ± 0.11
The voxel-based analysis also indicated increased [11C]raclopride binding in the ULD group (Fig. 2). The BP cluster in the right hemisphere partially covered the putamen and thalamus (2,546 voxels, peak voxel at [16 mm, −4 mm, −2 mm], corrected p < 0.001, significant also at FWE-corrected voxel-level p < 0.001). The BP cluster in the left hemisphere similarly covered parts of the putamen and thalamus (2,826 voxels, peak voxel at [−20 mm, 14 mm, 2 mm], corrected p < 0.001, significant also at voxel level p < 0.001). No significant results were detected in the opposite contrast (controls >ULD).
The oldest patient had moderate atrophy of the frontal lobes, parietal lobes, pons, medulla oblongata, cerebellum and vermis, whereas MRI images of brain structures in three other patients showed mild atrophy of various brain areas when compared with control subjects. All the patients had mild or moderate atrophy of the vermis; the youngest patient did not have any other noticeable brain atrophy. Despite the diffuse brain atrophy, the oldest patient had the highest [11C]raclopride BP in the thalamus, as presented in Fig. 1.
This is the first study to investigate dopamine receptor status in any of the diagnostic subgroups of PME, and only few studies have previously investigated in vivo the anatomical and functional involvement of the dopamine network in human epilepsy (Biraben et al., 2004; Bouilleret et al., 2005). Our results indicate that the BP of the D2-like receptor antagonist, [11C]raclopride, is dramatically higher in ULD patients than in age- and sex-matched healthy controls in the striatum and thalamus. Remarkably, there was no overlap of BPs between the studied groups, which made our result statistically highly significant despite the small number of patients. To our knowledge, such an extensive upregulation of [11C]raclopride binding in the striatum and thalamus has not been detected in any other disease or symptom. No differences were detected in the SUV data of cerebellar uptake in the patients versus controls for [11C]raclopride, thus supporting the use of the cerebellum in ULD as a reference region for nonspecific binding.
Increased BP of [11C]raclopride in the striatum and thalamus of the ULD patients can be due to decreased intrasynaptic dopamine levels and/or increased receptor density. Already in 1980, Leino et al. speculated that the decreased concentrations of homovanillic acid (HVA) in the cerebrospinal fluid (CSF) of Finnish PME patients versus healthy and epileptic controls indicate damage in dopaminergic neurons (Leino, 1980), and a common dopaminergic dysfunction in ULD appeared probable already over a decade ago (Mervaala et al., 1990). Previously, it has also been reported that one of two brothers suffering from dentatorubral-pallidoluysian atrophy, which is a subgroup of PME, showed increased [18F]dopa uptake in the striatum, and the same patient had spontaneous myoclonus unlike the other brother, who had decreased [18F]dopa uptake (Minami et al., 1994). As the thalamus responds to a large variety of sensory stimuli, including auditory, visual and somatosensory stimuli, it is conceivable that the increase in D2-like binding in the thalamus contributes to the development of the phenotype of sensory stimuli evoking myoclonus in ULD patients. It has been shown that a “fixed frequency” can trigger different responses including movement-activated myoclonus, stimulus-sensitive myoclonus, and normal muscle contractions (Panzica et al., 2003). This finding led to the suggestion that homogeneous neuronal pools endowed with distinctive firing frequencies drive both pathological and normal movements in ULD (Panzica et al., 2003). On the basis of our results, we suggest that these homogeneous neuronal pools are dopaminergic neurons in the striatum, and especially in the thalamus. The localization of the thalamic finding suggests primary effects in the intralaminar, midline thalamic, and anterior nuclei. It is possible that these findings are associated with functional alterations in thalamocortical or thalamostriatal projections. However, these issues should be investigated in more detail in future studies, because the spatial resolution of PET is insufficient for accurate separation of intrathalamic nuclei. The same limitation applies to mesencephalic regions, such as the ventral tegmental area and the substantia nigra pars compacta. Despite the indisputable dopaminergic alteration in ULD patients, nondopaminergic mechanisms are probably also involved in the genesis of myoclonus.
We cannot exclude the possibility that antiepileptic medications (sodium valproate, clonazepam, levetiracetam, piracetam, and topiramate) and/or increased D2-like receptor binding affinity had some effect on the tracer binding in our study. The latter option is unlikely, even though the mutational status of dopamine D2-receptors in ULD has not been studied. Because the effect of antiepileptic medication on myoclonus has been more or less marginal in our patients, as in previously described cases (Kyllerman et al., 1991), a significant medication-induced thalamostriatal dopaminergic modulation of this magnitude also seems unlikely. Furthermore, such a medication-induced modulation would mean that a heterogeneous group of antiepileptic drugs has similar increasing effects on [11C]raclopride binding. However, there is no evidence that piracetam, topiramate, or GABAergic antiepileptic drugs could increase D2-like dopamine receptor binding in the striatum and thalamus. Levetiracetam has been shown to have a beneficial effect on myoclonus in ULD (Kinrions et al., 2003; Magaudda et al., 2004). In animal models of Parkinson's disease, dopamine-induced dyskinesias have been significantly reduced with levetiracetam (Hill et al., 2003). Therefore, levetiracetam may counteract abnormal neuronal firing patterns, but the antimyoclonic effect might also, to some extent, be mediated by dopaminergic action.
In theory, the increased BP of [11C]raclopride in the ULD patients might be a secondary finding caused by an epileptogenetic modulation of dopaminergic pathways. If that is the case, the primary symptoms of the disease could be expected to be generalized tonic–clonic seizures and/or repetitive absence seizures, which would progress in the course of the disease, causing bilateral dopaminergic depletion associated with myoclonus. In this case, antiepileptic treatment should slow or cease the excitotoxic epileptic activity, which in turn should alleviate the progressive nature of the myoclonic jerks. However, in some cases, there is no evidence of epileptiformic EEG activity in the myoclonic stage of ULD (Kyllerman et al., 1991), generalized tonic–clonic seizures are totally absent in some patients, the effect of antiepileptic medication on myoclonus is moderate (Kyllerman et al., 1991), and the myoclonic jerks worsen over time in spite of any medication. Previous studies have shown a significant decrease in [18F]dopa uptake in the striatum and substantia nigra of epileptic patients with long-lasting and drug-resistant seizures (Biraben et al., 2004; Bouilleret et al., 2005), suggesting that these patients have a decreased dopamine neurotransmission in the basal ganglia. It can be speculated that the dopamine depletion in the basal ganglia of these patients could affect the seizure frequency and malignancy, but whether this is a cause or a consequence of seizures remains unclear. Because ULD patients do not usually have long-lasting and drug-resistant seizures like the above-mentioned epileptic patients, who, on the other hand, do not suffer from severe myoclonus, it is conceivable that the possible dopamine imbalance in these patients differs substantially.
In summary, our PET results suggest that there is almost complete dopamine depletion in the thalamostriatal area in ULD. Because of missing data on dopaminergic PET at the onset of the disease, it would be premature to suggest that ULD may be characterized more as a thalamostriatal movement disorder than epilepsy. However, the exceptionally high [11C]raclopride binding in the striatum and thalamus in ULD suggests that PET may be a supplemental follow-up/diagnostic procedure for ULD. It seems conceivable to test, e.g., D1/D2-specific dopamine receptor agonists, as treatment for myoclonus in ULD patients, so that the future data on the efficacy of these medications do not have to completely rely on enigmatic hypotheses.
Acknowledgment: We thank all the patients for their patience, help, and cooperation. We also wish to thank Professor Juhani Knuuti and the department secretary Mirja Jyrkinen from the PET centre for arranging convenient timetables. The study was financially supported by Turku University Hospital (EVO funds).