A workshop was held at National Institutes of Health in June 2005 to identify research priorities for understanding and treating spasmodic dysphonia (SD). The workshop was sponsored by the NIH Office of Rare Diseases, the National Institute on Deafness and Other Communication Disorders, the National Institute of Neurological Diseases and Stroke, the Movement Disorder Society, and the National Spasmodic Dysphonia Association. Fourteen specialists in otolaryngology, neurology, and neuroscience were invited to participate.
This summary is divided into four sections; each section presents a synopsis of workshop presentations followed by recommendations for research. The first section describes SD and lists recommendations for diagnosis. The second section summarizes what is known about the patient population and the types of studies that could help define the population better. In the third section, research and research recommendations to discover the causes and mechanisms in SD are discussed and in the fourth section, treatment options are discussed.
What is Spasmodic Dysphonia?
SD is a rare voice disorder that usually develops spontaneously in mid-life. Patients with SD have a hard time speaking; their speech may be strained or choked, or alternatively breathy, and patients all report that it requires a huge amount of effort to speak. After the initial onset, the disorder gradually progresses and then remains chronic for life. More women than men are affected; between 60 and 80% are female.
No diagnostic marker is available for SD; currently the speech symptoms are used and are difficult to differentiate from other voice disorders such as hyperfunctional voice or muscular tension dysphonia. The strained, effortful, intermittent voice that is characteristic of SD is due to involuntary spasms in the muscles of the larynx during speech. The spasms are intermittent and are task-specific, occurring only during speech. Adductor SD involves spasmodic hyperadduction of the vocal folds producing involuntary closing of the vocal folds, leading to voice breaks. In the less common form, abductor SD, involuntary opening of the vocal folds results in the prolongation of voiceless consonants as breathy breaks. Rarely both adductor and abductor spasms are present. Treatment outcome frequently serves as an important means to confirm the initial diagnosis; patients with hyperfunctional voice frequently resolve with voice therapy while patients with SD rarely benefit from voice therapy but are benefited by botulinum toxin injection. SD patients may have concurrent tremor, but tremor is rare with a hyperfunctional voice disorder. As the diagnosis is based on clinical characteristics, it often remains difficult to differentiate SD from hyperfunctional voice and other voice disorders. The following recommendations for the diagnosis of spasmodic dysphonia are an attempt at creating a first step in enhancing research on this disorder.
Recommendations for Diagnosis
The highest priority identified by the workshop is the need for a common diagnostic procedure. The workshop recommended a three-step process. The first step is a simple screening questionnaire that allows patients to identify themselves as possibly having SD, the second step is a clinical examination that identifies individuals with probable SD, and the third step requires a nasolaryngoscopy examination to confirm a definite SD. It was recommended that a multidisciplinary Spasmodic Dysphonia Study Group be established and charged with designing a validation study to determine the accuracy and reliability of the proposed diagnostic procedure.
Who gets Spasmodic Dysphonia?
SD is a rare disorder. It is also clinically heterogenous and requires experts to recognize and identify the disorder. These factors have made it difficult to get answers to basic questions regarding prevalence, incidence, age of onset, gender and ethnic representation, regional variation, and risk factors.
The estimated prevalence of SD is 1 case per 100,000 population. About 30% of patients with SD also have vocal tremor. No controlled studies of risk factors have been conducted; case series reports suggest an association of SD with autoimmune disease, thyroid disorders, or antihistamine use. Approximately 30% report a prior upper respiratory tract infection or suffering a major life stress (21%) at the time of onset.
Although 14 genes have recently been associated with dystonic disorders, none have been associated with SD. Genetic studies of SD have not been done because large families with only SD have not been described.
New technology is becoming available that will allow for the identification of genes from cohorts of unrelated patients and will make it possible to search for genetic association in a disease cohort compared to a control cohort. While powerful, this method requires a large number of patients with a standard clinical presentation as well as considerable sums of money. The impetus for finding a gene is that it then becomes possible to develop both animal and cellular models of the disease to study the biochemical cascade of events, to develop hypotheses, and then to try and intervene in the physiologically relevant pathways to halt the development of the disorder.
Recommendations for Population Research
Several of the following recommendations for population research were identified as a lower priority given the difficulties with accurate diagnosis. Implementation of these recommendations awaits the development of accurate diagnostic procedures.
1. Brain Banks
Several brain banks already exist and the SD community may be able to collaborate with existing banks. Annual exams need to be given to the living donors to rule out additional neurological diseases/disorders. The SD Study Group was charged with identifying and organizing sites for an SD brain bank.
2. Genetic Research
Going after a gene for SD was described as a million dollar experiment with no guarantees and at this time is considered premature. Once diagnostic criteria are standardized, evidence is needed regarding the assumption that a common etiology exists between SD and other focal dystonias. The first steps would be to standardize clinical criteria, to collect family histories, and to deposit a blood sample in a central public repository.
3. Patient Registry
The pitfalls of a patient registry were discussed; principal among them was the fact that a voluntary registry would not contain accurately diagnosed cases. A registry may be a useful way to identify patients who would be willing to participate in clinical trials [in which case the questions must be approved by an Institutional Review Board (IRB) and must adhere to Health Insurance Portability and Accountability Act (HIPAA) guidelines].
What are the Neurological Abnormalities in SD?
The mammalian vocalization system consists of a primitive, innate, preprogrammed system that produces laughing, crying, and other involuntary sounds as well as a second, more recently evolved, system for the control of learned vocalizations in humans. All subhuman mammalian species possess the innate system, while only humans have developed the system for learned vocalizations. This means that animal models for SD will not mimic the phenomenology of SD that affects voice production only during speech and not during laughing, crying, and other involuntary sounds.
Studies in other dystonias suggest there may be a deficiency in the caudate/putamen. Others suggest involvement of the dopaminergic pathways in dystonia, although the specific pathway and the type of defect still remain to be identified. Loss of inhibition is well documented in dystonia. Treatments for dystonia may increase inhibition, or replace inhibition that is missing. Increased neuroplasticity has been demonstrated in patients with writer’s cramp, the form of focal dystonia thought to be most similar to SD. In patients with SD, increased neuroplasticity could set the stage for a disease response to repetitive activity, injury, or increased use. Because the sensory system is a prime driver of the motor system, disordered sensory input might lead to disordered motor output.
In the laryngeal muscles, where the symptoms of SD are manifested, the emerging picture is one of disordered inhibition. The larynx is extremely sensitive to any stimulation of the mucosa. Receptors are present for not only tactile stimulation but also to negative pressure, air flow and cold, a lack of chloride ions, and muscle activity. Studies suggest sensorimotor disorders, such as reduced inhibition of motor responses to sensory feedback, may be present in SD.
Recommendations for Basic Research
1. Animal Models
An animal model may or may not mimic the clinical state of SD because no other mammals or primates exhibit learned vocalizations similar to humans. However, if the underlying brain disorder can be modeled, the model may be useful despite differences in laryngeal behavior across species.
Several animal models of generalized dystonias (mostly in rodents) are available; one involves striatal dopamine depletion and slight muscle weakening. Such a strategy may be applicable to the development of a model for SD. Songbirds might be considered as a model for SD. Vocalization by songbirds shares some similarities to human speech having both innate and learned sounds, although there are differences in the peripheral and central nervous systems involved in human speech and birdsong. If a suitable model were identified, experimental manipulations that could predict a cause or new treatments for SD could be performed.
2. Basic Brain Circuitry
Fully understanding the anatomy and circuitry in the larynx and the brain provides the foundation for further research. The anatomy of laryngeal reflexes, the specific connections between sensory relay stations, the motor inputs to the nucleus ambiguus, and descending connections from cortex to brainstem nuclei all require further research. This research would have to be done in animals, and the choice of animal was debated (rodents versus primates). As the circuitry is defined, higher brain centers that control and modulate laryngeal output need to be identified as well as drugs to modify these centers. The brain centers identified in animal studies could be compared to areas in human SD brains that appear abnormal from neuroimaging studies. In addition, neuropathological data from postmortem studies of SD brains should also be compared to the emerging animal data.
3. Related Experiments
a. Some evidence exists that an autoimmune event may be associated with development of SD. A study that could establish whether a link does/does not exist would be worthwhile.
b. Some patients with SD have had 70 injections of botulinum toxin into their larynx. The cumulative effects on the muscle are not known. Animal studies that reproduce multiple injections should help to understand the long-term effects of botulinum toxin injection on muscle physiology.
Developing Effective Treatments for SD
The most common treatment for SD is botulinum toxin injections into laryngeal muscles. Both botulinum neurotoxin type A and type B work by blocking release of acetylcholine at the neuromuscular junction and consequently reducing motor activity. About 90% of patients with adductor SD improve after receiving an injection for 3–12 months. Patients with abductor SD also respond to botulinum toxin injections, although it is not as effective and a slightly higher risk is involved since the muscles injected are also required for respiration. The disadvantage of botulinum toxin injection is that it must be repeated every 3–6 months. The dose response range is very wide; some patients require up to 20 times what other patients need to receive the same benefit. In addition, long-term results and safety have been challenges that still need to be addressed.
When surgical approaches to treatment of SD are considered, the possibilities include laryngoplasties, myomectomies, and denervation/reinnervation. Laryngoplasty involves changing the cartilagenous skeleton of the larynx, either through anterior commissure push back, thyroid cartilage widening, or medialization of the vocal cords. Patients who have these procedures usually do not experience lasting benefit. Myomectomies cut the muscles of the larynx, and the thyroarytenoid, lateral cricoarytenoid, and posterior cricoarytenoid muscles have all been cut. These procedures can be performed through a thyroid cartilage window or by endoscopic means. Dysfluency can be improved, although vocal harshness may result and lasting benefit has been unpredictable. The denervation/reinnervation procedure is currently being used in patients with pure adductor SD; patients with significant tremor or mixed dystonia are not candidates. The procedure is performed bilaterally. Women benefit more than men. Questionnaires returned by two-thirds of patients report some benefit. A small number went back to botulinum toxin injections, which are still possible after this type of surgery. Some patients reported having mild swallowing difficulties.
Recommendations for Treatment Research
Because placebo effects tend to be quite large in this patient group, researchers need to be wary. It is important to have controlled studies and check for previous treatments and concomitant signs and symptoms. An adequate follow-up duration of 3–5 years is important, particularly as SD is a chronic condition. Outcome measures should include subjective scales, objective tests, and quality of life questionnaires. Blinded videotape review is recommended where videotapes from before and after treatment are randomized and rated by blinded investigators. Small, open-label trials would be a good way to start discovering new treatment options. Since so little has been done, small trials would be a better use of resources than to commit to a big, randomized, placebo-controlled trial.
1. Confirm Physiology
Further imaging studies should be performed to confirm that the pathophysiology found in other dystonias is applicable to SD. Such studies could also help to define SD and place SD in context with other dystonias.
2. Pharmacological Treatments
Only the tremor component in SD modestly responds to oral medications such as beta blockers, although the SD tremor does not. Some patients receive a combination of botulinum toxin injections and a low dose of an oral, the combination possibly being more effective than drug monotherapy. These reports are anecdotal and prone to placebo effects. The recommendation is to systematically test combinations of botulinum toxin injections with a low dose of oral medications, testing known anticonvulsants, antidepressants, benzodiazepines, or dopamine agonists. Immunosuppressive agents might also be worth trying, especially in patients with incipient SD.
3. Surgical Procedures
Information should be collected on all patients who are receiving surgery for SD. The type of procedure should be documented with a standardized preoperative history and baseline and postoperative testing. Videotapes should be rated by blinded independent reviewers. Even an attempt to collect retrospective information from patients who have already received surgery is worthwhile, as well as collecting information on prospective patients. This approach, while not a clinical trial, might be valuable to catalog and disseminate information on surgical outcomes. A trial to evaluate the efficacy of various surgical procedures is greatly needed. Surgeons at multiple centers would need to be involved and trained to perform the procedure. Thorough long-term follow-up at multiple centers would be most convenient to patients and increase long-term information. Animal studies are needed to understand the long-term effects of surgery on laryngeal physiology.
4. Deep Brain Stimulation
Deep brain stimulation has proven to be effective in other neurologic conditions, such as Parkinson’s disease and essential tremor. However, it is unclear whether this will be effective for isolated SD and therefore caution should be used and systematic data collected.