Child Neurology: A young child with an undiagnosed case of dystonia responsive to l-dopa

Discussion

Levodopa was first used empirically for a hereditary childhood-onset dystonia with diurnal variation in 1971.¹ The patients were eventually shown to have deficiency of GTP cyclohydrolase 1 (GTP-CH-I), which is encoded by GCH1.² Monoamine neurotransmitter production depends on GTP-CH-I to occur as it is an essential cofactor. Deficiencies in other enzymes involved in the synthesis of dopamine have also been shown to respond to treatment with l-dopa, including tyrosine hydroxylase (TH), 6-pyruvoyl tetrahydrobiopterin synthase (PTPS), and sepiapterin reductase (SR). In addition to deficiencies in enzymes responsible for dopamine biosynthesis, other conditions, including spinocerebellar ataxia type 3 (SCA3), ataxia telangiectasia, hereditary spastic paraplegia type 11, parkin deletions, lesions of the substantia nigra, paroxysmal nonkinesigenic dyskinesia, and others, have been shown to involve dystonia that may respond incompletely to treatment with l-dopa. Further conditions that may partially respond to dopamine include glucose transporter type 1 deficiency, dihydropteridine reductase deficiency, aromatic L-amino acid decarboxylase deficiency, dopamine transporter deficiency, vesicular monoamine transporter 2 deficiency, SOX6 mutation, juvenile Parkinson disease, pallidopyramidal syndrome, and DYT1 mutation.^3,4

The idea of a diagnostic therapeutic trial with l-dopa has been traced back to Nygaard et al⁵ but more recently questioned by Maas et al.⁴ Over time, the debate on whether “every child with dystonia deserves a trial with l-dopa” has gone on. Some recent reviews argue for an empiric trial of levodopa.^3,6 Others argue against an empiric trial until a definitive diagnosis of DRD has been made for multiple reasons including the risk of side effects with l-dopa and the relative ease of diagnosis.⁴ Side effects from treatment with levodopa are possible and include headache, nausea, paresthesias, confusion, and dyskinesias; Lumsden et al.⁷ recently reported that 26/57 children treated with levodopa for dystonia eventually discontinued use due to side effects. In addition, the range of dosages required for successful treatment has wide variability, from 0.5 up to 20 mg/kg/d (although most children respond to 4–5 mg/kg/d).^4,8 Maas et al.⁴ further qualified their argument that patients with classical DRD could be given a therapeutic trial of l-dopa with concurrent further investigation with genetic testing. It could alternatively be argued, however, that even some cases of classical DRD do not present with diurnal variation. Furthermore, some patients with classical DRD respond rapidly to low-dose l-dopa, so a brief trial at low dose may be a reasonable first step in all young patients with dystonia in order to guide the next steps of evaluation.

Our patient did not present with classical DRD given the lack of diurnal fluctuation. However, we performed an empiric trial of l-dopa (rather than CSF neurotransmitter analysis) because it was not considered necessary to perform a procedure that would require sedation when the option for an alternative was available. Should the trial at low dose have been negative, we had the intention to reevaluate the need for further workup. Given the dramatic response to low-dose and once daily l-dopa administration, this was not needed.

Subsequent workup revealed white matter changes on MRI suggestive of delayed myelination vs dysmyelination. Given the otherwise benign course, delayed myelination was favored; however, repeat MRI 6 months later was unchanged, which suggested a dysmyelinating process. None of the classically cited DRDs (GCH1 deficiency and TH deficiency, in addition to less common conditions responding to l-dopa such as SR deficiency, PTPS deficiency, and SCA3) are associated with abnormal white matter. Therefore, adding in potential leukodystrophy/hypomyelination diseases on top of the already known l-dopa-responsive dystonia drastically increased the potential diagnoses. WES was performed, which was negative for any pathogenic variant or variant of unknown significance for any known causes of dystonia or myelination disorder. Interestingly, our patient's mother developed similar symptoms of dystonia with diurnal variation during the evaluation process, pointing towards a presumed genetic cause even with the negative WES.

At this point, a lumbar puncture for CSF neurotransmitters will not be performed, since exogenous administration of l-dopa may interfere with interpretation, and we do not believe that even temporarily discontinuing treatment for a diagnostic evaluation would be in the patient's best interest. WES has several limitations, including inability to detect copy number variants, analyze epigenetic factors, look for uniparental disomy, and analyze for nucleotide repeats, and not all exons can be tested completely. In addition, promoter regions, introns, and other noncoding regions are not analyzed in WES but could still contribute to disease processes. The patient will be referred for whole genome sequencing as the next step in his evaluation.

Proposed diagnostic strategies for dystonia that can reveal classic DRD include genetic testing for GCH1 and TH; further testing for less common causes of DRDs (SR and PTPS deficiency); SPECT imaging; lumbar puncture for neurotransmitter metabolites, neopterin, and tetrahydrobiopterin; and next generation sequencing/WES. However, the heterogeneity of other known conditions that can (more rarely) respond to treatment with l-dopa makes testing for every possible genetic cause of a potential l-dopa-responsive dystonia more difficult. Waiting for a firm diagnosis of our patient's dystonia would therefore have prevented or drastically delayed treatment with l-dopa, which was conclusively shown early in his course to completely correct his dystonic episodes.