The natural course of idiopathic cervical dystonia

Idiopathic cervical dystonia (ICD) is by far the largest subgroup of dystonia. Still, its natural course is largely unknown. We studied the natural course of 100 ICD patients from our botulinum toxin clinics (age at ICD onset 45.8 ± 13.5 years, female/male ratio 2.0) over a period of 17.5 ± 11.5 years with follow-ups during botulinum toxin therapy and with semi-structured interviews. Two courses of ICD could be distinguished by symptom development of more or less than 6 months. ICD-type 2 was less frequent (19% vs 81%, p < 0.001), had a more rapid onset (8.7 ± 8.0 weeks vs 3.8 ± 3.5 years), a higher remission rate (92% vs 5%, p < 0.001) and a higher prevalence of excessive psychological stress preceding ICD (63% vs 1%, p < 0.001). In both ICD-types, the plateau phase was non-progressive. Significant differences in patient age at ICD onset, latency and extent of remission, female/male ratio and prevalence of family history of dystonia could not be detected. ICD is a non-progressive disorder. ICD-type 1 represents the standard course. ICD-type 2 features rapid onset, preceding excessive psychological stress and a high remission rate. These findings will improve prognosis, treatment strategies and understanding of underlying disease mechanisms. They contradict the widespread fear of patients of a constant and continued decline of their condition. Excessive psychological stress may be an epigenetic factor triggering the manifestation of genetically predetermined dystonia.


Introduction
The course of a disease is important for physicians to plan therapies, for patients to anticipate their prognosis and for basic scientists to understand the underlying disease mechanisms.
Idiopathic dystonia is defined as dystonia occurring in the absence of identifiable inherited or acquired causes and in the absence of a psychogenic aetiology (Albanese et al. 2013).Idiopathic cervical dystonia (ICD) is the largest subgroup of dystonia (Dressler et al. 2022).For special genetic dystonias, such as DYT1 dystonia (Oppenheim dystonia), early onset progressive courses have been described (Klein and Muenchau 2013).However, only few patients fall in this category.For ICD, the natural course has not been described in detail (Pana andSaggu 2021, Van Zandijcke 1995).As dystonia is generally considered a neurodegenerative disorder, its course is usually perceived as being progressive, similar to Parkinson's disease and Alzheimer's dementia.This means for physicians, that treatment strategies may need to be escalated.For basic scientists, it means that certain underlying disease mechanisms may be assumed and others may be discarded.Most importantly, this means for patients, that they live under the fear of a constant decline of their condition, both with respect to severity and with respect to spread to other body parts.
This study wants to describe the natural course of ICD.Results presented here will considerably change the conventional perception of ICD being a progressive disorder.

Design
The study is based on regular follow-ups complemented by semi-structured interviews of ICD patients.Most of them received regular botulinum toxin therapy at the Movement Disorders Section of the Department of Neurology of Hannover Medical School in Hannover, Germany.

Patients
Patient inclusion criteria consisted of (1) diagnosis of ICD (2) Existence of ICD for at least one year (3) Willingness and ability to participate in the study (4) Regular follow-ups.Exclusion criteria included symptomatic cervical dystonia, psychogenic cervical dystonia and non-dystonic torticollis.Patients were seen in our botulinum toxin clinics and included consecutively into this study until the pre-set number of 100 patients was reached.

Study parameters
The study parameters are shown in Table 1.Patient demographics described the patient's sex, the patient's age at study inclusion and the patient's age at ICD onset.The patient's family history of dystonia derived from personal examination of family members, diagnoses by other physicians and signs and symptoms provided by the patient.The level of diagnostic accuracy may be best described as 'probable'.
The course of ICD is described by its onset phase, its plateau phase and its remission phase, as shown in Fig. 1.The observation period was defined as the time between ICD onset and last contact with the patient.In a semi-structured interview, the patient is shown Fig. 1 together with definitions of the onset phase, the plateau phase and the remission phase.Explanations were given in plain German language and the patients were allowed to ask clarifying questions until they and the interviewer thought the interview requests were understood.Based on the patient's information, the patient's individual ICD course was constructed.
Exceptional psychological stress was defined as psychological stress of a severity the patient had never experienced before or thereafter.

Follow-ups
Neurological follow-ups were mostly performed during the continued botulinum toxin injection series usually scheduled at intervals of 3 months.

Statistics
Study parameters were transferred into a matrix and correlations were calculated wherever they were of interest.The statistical tests applied are given in the text.All averages are given as mean ± standard deviation.Statistical significance was set to p ≤ 0.05.

Patients
Altogether 100 consecutive ICD patients were included in this study.There were no dropouts.The patient age at ICD onset was 45.8 ± 13.5 years.The patient age at study inclusion was not used for further evaluations, as it may be confounded by the presence or absence of remissions.66% of the patients were female, 34% were male, i.e. the female/male ratio was 2.0.The observation period was 17.5 ± 11.5 years.

ICD course
After preliminary screening of the onset phase duration, patients were divided into two distinct groups as shown in Table 2. Separation was based on an onset phase of more or less than 6 months.

ICD-type 2 (rapid onset)
19% of all ICD patients belonged to this group.Their age at ICD onset was 41.5 ± 12.7 years.63% of ICD-type 2 patients experienced excessive psychological stress.16% of them had a family history of dystonia, 63% of them were female, 37% male, i.e. the female/male ratio was 1.7.The natural course of each ICD-type 2 patient is shown in Fig. 3.The onset phase of ICD-type 2 patients was 8.7 ± 8.0 weeks (minimum 1 week, maximum of 24 weeks by definition).63% of ICD-type 2 patients experienced remissions.They started 1.2 ± 0.4 years after ICD onset with an extent of 52.5 ± 34.4% (minimum 10%, maximum 100%).As in ICDtype 1 patients, the plateau phase was never progressive.16% of ICD-type 2 patients developed mild additional dystonia manifestations after ICD onset including arm dystonia, spasmodic dysphonia and blepharospasm (no statistical analysis due to small sample sizes).One % of ICD-type 2 patients developed mild additional dystonia manifestations before ICD onset (no statistical analysis due to small sample size).None of these additional manifestations dominated the clinical picture.

Patients with excessive psychological stress
13% of all ICD patients experienced excessive psychological stress preceding ICD onset, including partner conflicts (divorce and separation, domestic violence), special familial burdens, legal disputes and migration.In patients with excessive psychological stress, age at ICD onset was 39.0 ± 13.9 years, onset phase 0.3 ± 0.8 years, remission rate 92%, remission extent 54.5 ± 35.3%, female/male ratio 1.2.

Botulinum toxin therapy
97% of all patients in this study received botulinum toxin therapy.The duration of botulinum toxin therapy was 11.5 ± 9.8 years (minimum 1 year, maximum 41 years).Efficacy of botulinum toxin therapy was a subjective improvement of 73.2 ± 13.9% (minimum 30%, maximum 90%).No patient developed antibody-induced therapy failure and no patient terminated botulinum toxin therapy.

General observations
With 100 patients included and an observation period of 17.5 ± 11.5 years, this study is-to our knowledge-the most detailed study on the natural course of ICD.

Patients
ICD patients reported here, confirm previous demographical features with a patient age of 63.6 ± 12.7 years, an age of ICD onset of 45.8 ± 13.5 years and a strong female sex preponderance of 2.0.

Botulinum toxin therapy
Botulinum toxin therapy is used in 98% of our patients.The subjective therapy effect indicates an improvement of 73.2 ± 13.9% including dystonic head movements as well as pain.During an observation period of 11.5 ± 9.8 years, we have not seen antibody-induced therapy failure, nor have we seen therapy termination.However, as this was not a prospective study, we have to refrain from definite statements.

Family history of dystonia
26% of all ICD patients had a family history of dystonia with a wide spectrum of dystonia manifestations, often different from the dystonia manifestation of the index case.This confirms, that dystonia manifestations within the families may vary considerably.As most of the family members could not be examined personally, the diagnosis had to be described as 'probable'.It is noteworthy, that reports on the family history dating back to more than one generation are usually not available.Currently, cases of tremor without other dystonic manifestation would not be diagnosed as dystonic tremor.Our study suggests, not to consider clinical findings alone, but also to consider the family history in patients with tremor.In the future, classification systems of tremor should allow considering tremors without other dystonic manifestation as dystonic, when the patients family history

ICD course
The natural course of ICD can be divided into two distinct types with the very different features shown in Table 2.

ICD-type 1 (regular onset)
With 81% of all ICD patients, this type covers the standard situation.ICD-type 1 develops over a period of 3.8 ± 3.5 years.Thereafter, its course is constant.Further progression would be uncommon, but might occur in special rare cases.In 26% of patients, mild additional dystonic manifestations may occur after ICD onset.Remissions are very rare, are mild to moderate at best and occur late, often when patients retire.This course resembles the course previously described (Meares 1971).It remains unclear, whether these remissions reflect actual improvement of the condition or increased resilience, because of improved resources after retirement.With a gradual onset over about 4 years and a constant course, ICD-type 1 is a non-progressive disorder clearly different from other neurodegenerative conditions such as Parkinson's disease and Alzheimer's dementia.

ICD-type 2 (rapid onset)
This type covers 19% of all ICD patients.ICD-type 2 develops rapidly within 8.7 ± 8.0 weeks.63% of these patients experienced excessive psychological stress preceding disease onset.For patients, this situation seems dramatic with its rapid onset and its often substantial disease severity.However, about two thirds of patients will experience substantial remissions within 1.2 ± 0.4 years.These remissions might be reflected in previous reports of ICD remissions (Bräutigam 1954;Jayne et al. 1984;Friedman & Fahn 1986).Very rarely, patients may experience repeated episodes of ICD-type 2. Patient age at ICD onset, patient sex, family history of dystonia and additional dystonia manifestations before or after ICD onset are not different between both types.The decisive difference between both ICD-types is preceding excessive psychological stress.Rapid disease onset in our patients with ICD and preceding excessive psychological stress should not be confused with abrupt disease onset in patients with psychological or functional disease.

Differential diagnosis of rapid onset torticollis
Rapid onset neck spasms, generally known as torticollis, are seen in the conditions featured in Table 4.Most commonly, rapid onset neck spasms may be caused by nondystonic mechanisms, typically by mechanical irritation of the peripheral neuromusculoskeletal system (Maigne et al. 2003).These cases of non-dystonic torticollis are typically tonic and painful.They recover quickly, especially under analgesics and muscle relaxants.
If rapid onset torticollis is caused by dystonia, it is probably most frequently caused by application of dopamine receptor blocking agents.Typically, young men may produce oromandibular lingual dystonia immediately after application of classical neuroleptics.ICD may also be part of such an acute neuroleptic reaction.Prompt response to anticholinergics is a hallmark of this condition.Rapid onset torticollis of dystonic origin may also be caused by sudden central nervous system pathologies such as stroke (LeDoux and Brady 2003).These cases are extremely rare and usually have a static course.
Recently, psychogenic or functional dystonia was defined as a manifestation of dystonia, rather than pseudo-dystonia as before.Abrupt onset-often in conflict situations-is a hallmark of psychogenic or functional movement disorders (Hallett 2016).Abrupt onset in psychogenic movement disorders is typically within hours or days, thus different from our ICD-type 2 patients with rapid, but not abrupt onset.Valid data about prevalence, natural course and treatment of psychogenic ICD are still lacking.A recent study suggests a prevalence of psychogenic dystonia of 28.5/1mio or of 5% of all dystonia patients (Dressler et al. 2022).Within psychogenic dystonia, psychogenic ICD is relatively common.In our study, non-dystonic torticollis, drug-induced ICD, other symptomatic ICD and psychogenic ICD were explicitly excluded.

Excessive psychological stress
13% of all ICD patients experienced excessive psychological stress preceding ICD onset.In 85% of these patients, ICD developed within 5.8 ± 4.4 weeks, then lasted 18.5 ± 8.3 months, before it started to remit after 2.7 ± 0.8 years to 54.5 ± 35.3% of its maximal severity.ICD with preceding excessive psychological stress is closely connected with ICD-type 2: 63% of patients with ICD-type 2 experienced preceding excessive psychological stress and 84% of patients with preceding excessive psychological stress presented with ICD type 2. It is challenging, to elucidate the interactions between ICD and excessive psychological stress.This was performed in another publication of ours (Dressler et al. 2023).

Patient counselling
Our data clearly show, that both ICD-types are non-progressive.For patient counselling, we assume no major further deterioration, once ICD has been stable for two or three years.This is important information for patients, as most of them intuitively assume a chronic progressive course similar to other neurodegenerative disorders such as Parkinson's disease or Alzheimer disease.

Outlook
This is the first description of two distinct types of ICD.
For both types, the natural course is non-progressive.This distinction should be included in the classification of ICD, as it has immediate relevance for the counselling and treatment of our patients and for the basic scientist's understanding of the pathological mechanisms underlying ICD.Further studies will have to show, whether the findings described here for ICD will also apply to other focal and non-focal dystonias.
Funding Open Access funding enabled and organized by Projekt DEAL.

Fig. 1
Fig. 1 Schematic time course of idiopathic cervical dystonia.Explanation in text

Fig. 2
Fig. 2 The natural course of cervical dystonia-type 1 as reconstructed from data from 81 patients

Table 1
Study parameters with their definitions and dimensionsBT botulinum toxin, ICD cervical dystonia

Table 2
Comparison of patients with ICD-type 1 (regular onset) and ICD-type 2 (rapid onset) n/a not applicable, ns not significant, nsa no statistical analysis due to small sample sizes, w weeks, y years, ICD cervical dystonia

Table 3
Dystonia manifestations and their frequencies in 42 family members with history of dystonia High numbers of family members affected by tremor and a substantial proportion of patients with presumed essential tremor actually suffering from dystonia support this view.A family history of dystonia was not correlated with patient age, age at ICD onset, patient sex, rapid dystonia onset and remissions.

Table 4
Causes for rapid onset neck spasms