Dopa-Responsive Dystonia Revisited: Diagnostic Delay, Residual Signs, and Nonmotor Signs

Objective To investigate the delay in diagnosis, residual motor signs, and nonmotor signs of dopa-responsive dystonia (DRD) using literature and our own pilot data.

Design, Setting, and Patients We searched the MEDLINE database for patients with clinically typical DRD and/or guanosine triphosphate cyclohydrolase I gene mutations from 1952 to 2011 and examined a pilot cohort of 23 outpatients with DRD and guanosine triphosphate cyclohydrolase I gene mutations.

Results The literature search yielded 101 reports describing 576 cases. Excluding cases without proven guanosine triphosphate cyclohydrolase I gene mutations as well as homozygous and asymptomatic mutation carriers resulted in 352 cases. The mean (SD) ages at onset were 11.6 (13.4) years (literature) and 9.4 (7.7) years (pilot study). The average (SD) delays in diagnosis were 13.5 (13.3) years (literature) and 15.5 (16.3) years (pilot study); using all literature cases, they were 9.1 (7.5) years before and 15.2 (13.7) years after identification of the guanosine triphosphate cyclohydrolase I gene. Residual motor signs in patients receiving therapy were found in 28% (literature) and 39% (pilot study). Residual motor signs in the literature comprised dystonic (20%) and parkinsonian (11%) symptoms, as well as complications such as contractures or unnecessary surgical procedures. Information on nonmotor signs was given for 70 patients in the literature. Of these, 34% had depression, 19% anxiety, and 9% obsessive-compulsive disorder. Six of our own cases (32%) reported 1 or more nonmotor signs including depression and migraine.

Conclusions The delay in diagnosis is long, despite the well-known etiology and availability of genetic testing and specific therapy. A sizable number of treated patients have residual motor signs, nonmotor signs, and complications resulting from the lack of timely therapy or unnecessary procedures.

Dopa-responsive dystonia (DRD), a childhood-onset or adolescent-onset form of dystonia with marked diurnal fluctuation and an excellent response to levodopa, is among the best-described forms of monogenic dystonia.^1,2 The most common cause of DRD (ie, autosomal dominant mutations in the guanosine triphosphate cyclohydrolase I [GCH1 ] gene) was identified as early as 1994.³ However, there remains a marked delay in establishing the diagnosis even today, with far-reaching consequences on the disability of patients.⁴

The GCH1 gene encodes the rate-limiting enzyme for the synthesis of tetrahydrobiopterin, an essential cofactor for the conversion of tyrosine to levodopa,⁵ explaining the striking therapeutic response. However, many patients retain residual motor signs (RMS), the expression and extent of which have not been formally assessed to date. Even less attention has been paid to the potential clinical consequences of the fact that tetrahydrobiopterin is also involved in the biogenesis of serotonin, an important neurotransmitter in the regulation of mood and anxiety. Nonmotor signs (NMS) are increasingly recognized as an important feature of different movement disorders, such as Parkinson disease, and impact significantly on patients' quality of life.⁶ Despite the biological plausibility of NMS as a potential feature of DRD, to our knowledge, this has not yet been systematically studied and, along with diagnostic delay and RMS, is the third focus of our article.

We searched the online MEDLINE database (http://www.ncbi.nlm.nih.gov/pubmed) from the first description of the disease to December 2011, using the search terms dopa-responsive dystonia, DYT5, Segawa syndrome, GTP cyclohydrolase 1, and GCH1. This resulted in 1094 citations that were screened for information on clinical characteristics and the mutational status of probands. All studies providing individual data and, for NMS, 2 additional publications with group data were included. Families reported in multiple publications were excluded. We identified 576 probands with either a reported heterozygous GCH1 mutation or a typical clinical description of DRD. Excluding asymptomatic, compound heterozygous, and homozygous probands yielded a sample size of 492 patients with DRD (140 clinically identified cases and 352 genetically proven heterozygous-affected mutation carriers; Figure 1). Because the 140 clinically identified cases may have differed from the cases with a genetically confirmed diagnosis, we used the 140 cases for the analysis of the delay in diagnosis only. The 352 mutation carriers comprised 166 index patients and 186 affected family members.

We collected information on the age at onset, the age at diagnosis, RMS while taking levodopa treatment, and NMS. Age at onset was defined as the age for which patients recalled the appearance of first symptoms. Age at diagnosis was defined as the time of first diagnosis and, if this was not reported, we used age at first treatment. We defined delay in diagnosis as the difference between the age at onset and the age at diagnosis. Whenever the authors reported the persistence of motor signs of any severity in the neurologic examination, such as parkinsonism or dystonia despite dopaminergic therapy (regardless of doses and duration of treatment), we considered those as RMS. The term complications was used to summarize contractures, secondary orthopedic deformities such as scoliosis, and unnecessary surgical procedures. For NMS, we aimed to collate all reported nonmotor features. However, only migraine and psychiatric features were mentioned in the literature; therefore, we focused on psychiatric features in all pilot examinations. A complete list of the reviewed articles (n = 101) is available in the eAppendix.

All 23 patients with DRD included in the pilot study had proven GCH1 mutations and were diagnosed or followed up at our Movement Disorder Center since 1995. The study was approved by the university's ethics committee, and all subjects gave written informed consent. Of those, 10 participated in a personal and standardized videotaped neurologic examination by movement disorder specialists (V.T., N.B., J.H., and C.K.) and an extended questionnaire survey. Dystonic and parkinsonian features were assessed with the Burke-Fahn-Marsden Dystonia Rating Scale and the Unified Parkinson's Disease Rating Scale. Information on the remaining 13 patients was obtained by self-report (mailed questionnaires) and additional medical records (3 cases). To ascertain current depression, we used the Beck Depression Inventory. In addition, lifetime occurrence of depression and a general question for other disorders were included.

Most data were summarized at a descriptive level. Otherwise, only univariate statistics were applied using χ² tests for categorical variables and t tests or analysis of variance for continuous variables.

Comparing patients reported before the genetic test became available (between 1952 and 1993) with patients reported from 1994 onwards, delay in diagnosis was longer after the availability of genetic testing, with a mean (SD) delay in diagnosis before 1994 of 9.1 (7.5) years and after 1994 of 15.2 (13.7) years.

Dystonia started more frequently in the lower limbs and less frequently in the upper limbs and neck in the early-onset cases (0-14 years) as compared with the late-onset cases (>15 years). In the latter cases, parkinsonism was more common than in early-onset cases. Women had younger age at onset (10.1 years) than men (15.2 years; P = .04), and women had more frequent onset of dystonia in the lower limbs (P < .001) (Figure 2).

For analyses of RMS, availability of information was incomplete (Table 1 and Table 2). Residual motor signs included dystonia or parkinsonism, and both were more frequently observed in men. Similarly, RMS were more frequent in younger-onset cases (Table 1). The frequency of complications did not differ by sex, but comparison by age revealed the presence of complications in the 2 youngest-onset groups only. The rate of complications was lower (33%) in the delay in diagnosis of less than 5 years group and 100% in cases with a delay in diagnosis of 5 years. Descriptions of NMS were scarce. Therefore, we included 2 additional reports not providing data on individual patients, resulting in 70 cases with information on NMS. Frequency of depression was 34%, 19% for anxiety, and 9% for obsessive-compulsive disorder (Table 2).

Our 23 patients (17 women and 6 men) comprised 12 index patients (10 familial and 2 sporadic) and 11 affected relatives. Mean (SD) ages were 9.4 (7.7) years at symptom onset, 23.2 (17.2) years at diagnosis, and 15.5 (16.3) years in the delay in diagnosis. Residual motor signs were reported by 52.9% of the women and none of the men. In a subanalysis of the examined 10 patients only, the frequency of RMS was 60%, with only women affected (residual dystonic signs, 66.7%; residual signs for parkinsonism, 22.2%). Six cases (32%) reported 1 or more NMS including depression, anxiety, and migraine. As an indicator of current depression, the Beck Depression Inventory showed mild or moderate depressive symptoms in 9 of 14 women (64.3%) and none of the men (Table 3). All patients in the pilot study were treated with dopaminergic medication and reported their NMS while receiving treatment.

Our study demonstrates that DRD is far from being the well-recognized condition one would expect it to be based on its pathognomonic clinical features, availability of conclusive biochemical and genetic testing, and excellent treatment options. Our finding of a considerable delay in diagnosis, both in the cases from the literature and our own series, has a high imperative, as DRD is one of the few neurogenetic disorders and the only form of dystonia for which a miracle healing can be achieved for the lifetime of the patient. In this context, it should be noted that we were obviously unable to assess the rate of undiagnosed DRD cases, which is likely high. Surprisingly, the availability of genetic testing after 1994 did not improve the delay in diagnosis; rather, although not a significant difference, the average delay in diagnosis rose from 9 to 15 years since 1994. However, these results have to be interpreted with caution, as several biases are likely. First, the summarized study types did not include population-based studies and in clinic samples, the risk for selection bias was particularly high. Additional selection bias was introduced via the large proportion of missing data; it is likely that the cases with complete information differed from those without. Second, a reporting bias emphasizing atypical or severe cases is conceivable, as indicated by the small number of cases reported to be free of a particular DRD symptom and the high frequency of RMS.

Although information on complications related to delay in diagnosis was explicitly given for only 23 patients, as expected, the delay in diagnosis seems to impact the risk for complications. In our pilot series, 1 patient had undergone 9 surgical orthopedic procedures of her feet, resulting in severe, permanent (nondystonic) impairment of gait. From a clinical point of view, it seems reasonable to focus special attention even on rare symptoms if they are severe and avoidable, as is the case for some of the DRD complications.

For analyses of clinical symptoms, the major challenge is the availability of systematic information. However, we were able to confirm the early onset of dystonia with a median of 8 years and the previously described spatiotemporal gradient of the distribution of dystonic signs in relation to age at onset. In addition, parkinsonism was more common in older-onset cases.

Residual motor signs were reported more frequently than described previously. These also included subtle motor signs that did not necessarily interfere with the patients' activities of daily life but were easily detectable on neurologic examination.

Information on NMS was incomplete but indicated high frequencies of depression, anxiety, and obsessive-compulsive disorder as compared with the general population, as well as to patients with Parkinson disease.^7-9 Interestingly, these 3 NMS are all linked to serotoninergic pathways highlighting their biological plausibility within DRD. From other diseases, such as Parkinson disease, it is well known that NMS may have a large impact on the quality of life.¹⁰

In conclusion, the remaining delay in diagnosis, complications resulting from a lack of timely therapy or unnecessary procedures, and the occurrence of RMS and NMS warrant further clinical studies of DRD. In addition, our study should alert clinicians to the importance of not only establishing an early diagnosis but also of continuously monitoring patients with DRD on dopaminergic medication for potentially treatable RMS and NMS.

Correspondence: Christine Klein, MD, Section of Clinical and Molecular Neurogenetics, Department of Neurology, University of Lübeck, Ratzeburger Allee 160, 23538 Lübeck, Germany (christine.klein@neuro.uni-luebeck.de).

Accepted for Publication: March 5, 2012.

Published Online: September 17, 2012. doi:10.1001/archneurol.2012.574

Author Contributions:Study concept and design: Tadic, Brüggemann, and Klein. Acquisition of data: Tadic, Brüggemann, Stiller, Hagenah, and Klein. Analysis and interpretation of data: Tadic, Kasten, Brüggemann, and Klein. Drafting of the manuscript: Tadic and Kasten. Critical revision of the manuscript for important intellectual content: Tadic, Kasten, Brüggemann, Stiller, Hagenah, and Klein. Statistical analysis: Tadic and Kasten. Obtained funding: Klein. Administrative, technical, and material support: Brüggemann, Stiller, and Klein. Study supervision: Hagenah and Klein.

Role of the Sponsor: The funders had no role in the design and conduct of the study; in the collection, analysis, and interpretation of the data; or in the preparation, review, or approval of the manuscript.

Additional Contributions: We thank the patients for their time and effort.