6.1 Introduction

Spasmodic dysphonia is a clinical syndrome characterized by involuntary, hyperfunctional spasms or postures of the intrinsic laryngeal musculature, yielding abnormal speech. Traube ¹ first used the term spastic dysphonia to describe patients with nervous hoarseness in 1871. Historically, the terms mogiphonia, spastic dysphonia, aspartic aphonia, phonic laryngeal spasm, and coordinated laryngeal spasm have been used to describe the same clinical presentation. ^{2 ,​ 3} Long thought to be psychological, spasmodic dysphonia was shown to be clearly organic and neurologic in nature by the dramatic response to nerve section in pioneering work by Dedo and Behlau. ⁴ In 1982, Marsden and Sheehy ⁵ proposed that spasmodic dysphonia represented a laryngeal dystonia. They noted that “all evidence points to the conclusion that blepharospasm and oromandibular dystonia seen in Meige [syndrome] is another manifestation of adult onset torsion dystonia, [and] since dysphonia may occur in the same syndrome, it is quite likely that dysphonia itself may be the sole manifestation of dystonia.” In 1988, Blitzer and colleagues ⁶ showed through clinical examination and laryngeal electromyography (EMG) that features of spastic dysphonia were entirely consistent with focal cranial dystonia. With growing advances in brain mapping and neurogenetics, the neurological underpinnings of spasmodic dysphonia and other laryngeal dystonias are being discovered. ⁷ Functional magnetic resonance imaging (MRI) has showed changes in the corticobulbar tracts connecting motor sensory centers within the cortex to phonatory motoneurons in the brainstem ⁸ and genetic screening continues to provide connections between laryngeal dystonias and other neurological diseases.

6.2 Classification and Presentation

Spasmodic dysphonia is classified as a laryngeal dystonia, a clinical term used to describe an action-induced, task-specific, laryngeal neuromuscular disorder resulting in disturbance in the production of connected speech. ⁹ Dystonias are classified by clinical symptom, age at onset, distribution, and cause. When classified by distribution, dystonias are categorized as focal, segmental, multifocal, or generalized. Spasmodic dysphonia is a focal dystonia involving the laryngeal adductor muscles (lateral cricoarytenoid, interarytenoid, thyroarytenoid, and possibly the cricothyroid), abductor muscles (posterior cricoarytenoid [PCA]), or both, with occasional involvement of supraglottic structures. The degree and topography of laryngeal muscle involvement appears to vary, and the clinical manifestation may actually represent the predominant activity rather than pure, mutually exclusive adductor or abductor muscle involvement. ¹⁰ Laryngeal dystonia may also affect other functions of the larynx besides connected speech. Both singer’s dystonia and respiratory adductor dystonia have been described. ^{11 ,​ 12}

The majority of dystonias are primary, or of idiopathic etiology, although as many as a third of patients identify a precipitating factor or incident. ¹³ These patients have a normal perinatal and early developmental history; no prior history of head trauma or neurologic illness; no exposure to drugs known to cause acquired dystonia (e.g., phenothiazines); and normal intellectual, pyramidal, cerebellar, and sensory examinations. Identification of any cause by such history, examination, or laboratory studies defines secondary dystonia. Most cases of spasmodic dysphonia have no manifestations outside of the larynx, consistent with their identifications a focal dystonias. Rarely, cases may be associated with or progress to symptoms in another part of the body. Although in our experience this is a very small percentage, rates as high as 17% have been reported. ⁷ Multiple factors have been noted to cause secondary laryngeal dystonia. These may include neurologic disorders, drug exposures, and Parkinsonism.

Clinically, spasmodic dysphonia is divided into abductor, adductor dysphonia, “singer’s dysphonia,” and adductor respiratory dystonia. In adductor spasmodic dysphonia, patients present with a choked, strain-strangled voice, breaks in phonation, diminished volume, and a monotonal pitch. Patients with the abductor type present with a breathy, effortful voice, abrupt breaks in fluency, and whispered elements of speech. With severe spasms, patients may be aphonic. Compensatory behaviors may mask the patient’s true voice pattern. Patients with a severe adductor spasmodic dysphonia may exhibit compensatory abductor voicing, with aphonia or whispering. ¹⁴ Although some authors believe that all patients with spasmodic dysphonia have abductor and adductor involvement, with symptoms manifesting according to the predominant type, ^{15 ,​ 16} the main premise of botulinum toxin treatment, chemodenervation of the hyperfunctional muscle group, is well served by the adductor–abductor classification.

In both groups, patients may demonstrate a phonation-associated tremor. In contrast to essential tremor, the spasmodic dysphonia–associated tremor is irregular and may be secondary to posturing of dystonic muscles in a position in which the agonist contractions do not fully neutralize those of the antagonist muscles. Blitzer and colleagues ¹⁷ showed that 25% of spasmodic dysphonia patients may have such a tremor. Distinguishing dystonic tremor from essential voice tremor may occasionally be exceptionally challenging.

6.3 Diagnosis and Investigation

The diagnosis of spasmodic dysphonia is clinical, based principally on perceptual analysis of the voice, complemented by the laryngoscopic examination. Evaluation should include a detailed head and neck and neurologic examination, with particular attention paid to spasm, dysfunction, or tremor in any area of the head and neck. Laryngoscopy should be performed with flexible transnasal instrumentation, because it disrupts physiologic laryngeal function least, and allows assessment of connected speech as well as swallowing and breathing. The rigid telescope may have a role for the detailed evaluation of the vibratory margin of the vocal fold, and to exclude mucosal wave pathology, but is likely to obscure the tell-tale signs of spasmodic dysphonia. The nuances of diagnosis are beyond the scope of this chapter, but the primary diagnostic challenge is differentiating between spasmodic dysphonia, muscle tension dysphonia (MTD), and vocal tremor, or combinations of the three. Leonard and Kendall ¹⁸ noted that motion abnormalities in spasmodic dysphonia were present only during specific speaking actions, whereas they were present more consistently during MTD and vocal tremor.

Identification of intermittent hyperfunctional breaks and spasms, with excessive and inappropriate adduction or abduction of the vocal folds during specific vocal tasks, is the basis for the diagnosis of spasmodic dysphonia. The spasms and breaks have a characteristic and reproducible pattern that can be elicited using a standardized set of sentences. During the endoscopic laryngeal exam, adductor-type spasmodic dysphonia typically demonstrates intermittent, occasionally more sustained, hyperadductory postures with hyperfunctional closure of at least the true and sometimes the false vocal folds, excessive medial rotation of the vocalis process, or excessive vocal fold tension that corresponds to the choked, strained-strangled voice breaks. Adductor spasms typically occur with voiced onsets. The laryngeal exam for abductor-type spasmodic dysphonia should exhibit inappropriate abduction during connected speech, which results in a breathy, although effortful, voice quality with aphonic or whispered segments. Abductor spasms are marked when trying to phonate a vowel after a voiceless consonant (example /h/, /p/, or /t/), in contrast to other causes of breathy voice from glottic insufficiency such as vocal fold paralysis or presbylarynges, which yield a more consistent breathy quality to the voice. The presence of secondary functional or compensatory postures may render the diagnosis difficult, and a short course of voice therapy may be useful to identify and address these.

Ludlow and colleagues emphasized the need for a standardized diagnostic criterion to help differentiate between vocal tremor, MTD, and spasmodic dysphonia ¹⁹ and proposed a three-tier system for the diagnosis of spasmodic dysphonia. Tier one is a screening questionnaire that consists of four questions (Table 6‑1).

Patients are expected to have had symptoms for at least 3 months and answer in the affirmative to the first two questions to be considered as having possible spasmodic dysphonia. Tier two is a clinical speech examination in which patients are taken through a series of vocal exercises by a voice specialist. Specific sentences are used to elicit abductor and adductor voice breaks. Patients are expected to have one or more voice breaks during speaking and fewer during whispering to be included. Tier three, fiberoptic laryngoscopy, reveals normal glottal function with swallowing, whistling, and coughing while demonstrating vocal fold spasm, or tremor while speaking sentences. Using this three-tiered technique on a series of 30 patients with known MTD, spasmodic dysphonia, or tremor, Ludlow and colleagues ²⁰ were able to correctly categorize 97% of patients.

Objective testing and other diagnostic modalities can be useful in the diagnosis of difficult cases. In assessing the utility of acoustic analysis, Zwirner and coworkers ²¹ found significantly higher mean values of standard deviation of fundamental frequency, jitter, shimmer, and voice break factor, and significantly lower mean values of signal-to-noise ratio in spasmodic dysphonia patients when compared with normal controls. Using acoustic analysis, Sapienza and coworkers ²² found that only patients with adductor spasmodic dysphonia manifested voice breaks during speech, specifically with sustained vowels. They also found patients with adductor spasmodic dysphonia to have greater variation in the type of acoustic event produced as a function of speech task. Koufman ²³ found spectral analysis to be useful in differentiating adductor spasmodic dysphonia and MTD, sometimes a challenging task given the supraglottic hyperfunction that may be present in both processes. Koufman noted that voice breaks were usually present in spasmodic dysphonia but absent in MTD. The spasmodic dysphonia patients also had well-defined formants, whereas the MTD patients did not, and the MTD patients had excessive high-frequency spectral noise that was minimal in spasmodic dysphonia patients. ²³ This showed that spectral analysis, although not a frequently used tool, may be a useful adjunct to differentiate adductor spasmodic dysphonia from MTD.

Typically, laryngeal EMG demonstrates abnormal, but not pathognomonic, findings. Large and polyphasic motor unit potentials and an abnormally long latency from the initiation of electrical signal to the beginning of sound production have been reported. ^{24 ,​ 25} Nash and Ludlow ^{26 ,​ 27} compared laryngeal EMG results in patients with adductor spasmodic dysphonia against controls. They found a significant increase in mean muscle activity during voice breaks in the thyroarytenoid muscle, unsurprising given the phenomenology of the disorder, whereas muscle tone in these patients was equivalent to controls during normal speech. Hillel ¹⁰ found abnormal laryngeal EMG patterns of response, increased latencies, and increased amplitudes of recruitment in many tasks including nonphonatory tasks among all spasmodic dysphonia patients. Most importantly, his work revealed demonstrated abnormal activity in all five intrinsic laryngeal muscles and previously undocumented variability in muscle involvement, revealing more complexity in the disease than commonly assumed.

About 12 to 15% of patients with spasmodic dysphonia have a positive family history and genetic testing shows promise as a diagnostic modality. Several genes have been associated with laryngeal dystonias. Notable gene regions and mutations that have laryngeal dystonia as a manifestation include TOR1A (DYT1), TAF1 (DYT4), THAP 1 (DYT6), and GNAL (DYT25). Most of these genes are associated with regional manifestations of dystonia or other neurological features such as Parkinsonism. At this point, however, routine genetic testing for isolated laryngeal dystonia has a low yield, but in the context of other neurological symptoms family history can be considered. ²⁸

Functional MRI features for laryngeal dystonia are quickly evolving and alterations in brain connections are being described with more specificity in spasmodic dysphonia. Functional MRI can show metabolic activity during tasks such as vocalization. Changes in the connectivity between different regions associated with voice tasks have been demonstrated in patients with spasmodic dysphonia. ²⁹

The diagnosis of spasmodic dysphonia can be made reliably with the patient’s history and physical examination. Acoustic analysis and laryngeal EMG both have shown promise as aids for difficult cases, but more research is required to better define their role. The cornerstone of diagnosis at this time remains a carefully directed history with appropriate screening questions, speech examination, and flexible laryngoscopy.

6.4 Treatment

Botulinum neurotoxin type A injections into the intrinsic laryngeal muscles represent the current standard for the treatment of spasmodic dysphonia. There have been hundreds of peer-reviewed articles investigating the utility of BoNT injections for spasmodic dysphonia, with the collective evidence overwhelmingly favoring the effectiveness of this modality.

In 2015, Blitzer et al updated their experience of BoNT treatment for spasmodic dysphonia. ³⁰ They found that 90% of patients improved for 3 to 12 months after injection of BoNT-A, with a need for repeat injections every 3 to 6 months. Two meta-analyses have been performed that considered the efficacy of BoNT, and one blinded, randomized controlled trial has been conducted in which BoNT injection was compared with saline injection and measured with objective acoustic outcomes. ^{31 ,​ 32 ,​ 33} BoNT-A markedly reduced perturbation, decreased fundamental frequency range, and improved spectrographic characteristics.

Although the main effect of BoNT is a blockade of the injected muscle at the neuromuscular junction in the peripheral nervous system, this explanation does not seem to correlate with the pathophysiologic model of spasmodic dysphonia or the surprising efficacy of this injection, suggesting an effect on the central nervous system. Byrnes and colleagues ³⁴ showed that BoNT injections caused a transient change in the mapping of muscle representation areas in the motor cortex. This has further been suggested by the effect of BoNT on noninjected laryngeal muscles in spasmodic dysphonia. ³⁵ In 2007, Antonucci et al ³⁶ demonstrated that BoNT is translocated to the afferent synapses in the contralateral hemisphere in mice and rats. Although these mechanisms are not completely understood, the efficacy and safety of BoNT have led to its being deemed the primary therapy for spasmodic dysphonia.

The target muscle for BoNT injection depends on the relative adductor and abductor features of the spasmodic dysphonia. The standard treatment for adductor spasmodic dysphonia at the majority of voice centers is bilateral EMG-guided transcutaneous injections into the thyroarytenoid muscle, using equal amounts of BoNT; however, there are several variations of this approach. The dose of BoNT given, the time course between injections, unilateral versus bilateral injections, and the follow-up can vary between centers and among individual practitioners. In addition, there seems to be great variability in the BoNT sensitivity and recovery among patients as well as fluctuations in the disease symptoms. In essence, each patient requires an individualized plan of care, and practitioners may need to alter their injection methods, doses, and schedules accordingly to maximize efficacy.

The efficacy of bilateral thyroarytenoid injections has been the most studied and has the most robust and longest treatment history. However, some small studies have shown that unilateral injections may have equivalent improvement in voice symptoms and sometimes improved therapeutic profiles. ^{37 ,​ 38 ,​ 39} Adams and colleagues ³⁷ compared 15 patients receiving a 15-unit (U) unilateral injection of BoNT with 11 patients receiving bilateral 2.5-U injections. They found that both the unilateral and bilateral BoNT injections were associated with significant improvements in spasmodic dysphonia, and both types of injections were associated with a significant increase in vocal breathiness at 2 weeks postinjection. However, in using acoustic analysis to compare the two, they found that maximum phonation time, vocal jitter, and the number of voice breaks per second indicated that unilateral BoNT injections may provide superior and longer lasting benefits than bilateral BoNT injections. The same group again compared unilateral and bilateral injections in 1995, treating 25 patients with a 15-U unilateral injection and 25 patients with bilateral 2.5-U injections. They found comparable results at 2 and 6 weeks using acoustic analysis. However, the bilateral group was noted to have a significant reduction in maximum phonation time in comparison with the unilateral group. ³⁸ Upile and colleagues ³⁹ examined 31 patients with adductor spasmodic dysphonia who received either unilateral or bilateral injections. They found that bilateral injections were associated with postinjection voice loss, whereas unilateral injections were not.

Ford et al ⁴⁰ described indirect laryngoscopy for guidance in performing thyroarytenoid injections. In their study, a slightly delayed time of onset was noted. However, the efficacy and duration of the injections were not significantly changed from the standard EMG-guided injections. They pointed out that although most otolaryngologists are unfamiliar with laryngeal EMG, most are comfortable with laryngoscopy. Flexible laryngoscopy has been used both to visualize percutaneous thyroarytenoid injection and to guide injections through the channel of the laryngoscope. ^{41 ,​ 42 ,​ 43} Although perhaps more familiar, these approaches have limitations. Transoral injection may be limited or made impossible by the patient’s gag reflex. Injection through a flexible bronchoscopic needle entails waste of BoNT to fill the needle tubing and makes precision difficult to control. ⁹ The aforementioned techniques also do not allow for EMG confirmation of needle placement, which allows for controlled administration of treatment into the more actively contracting regions of the muscle, near motor end-plates and the site of BoNT action.

Injection of the thyroarytenoid muscles may be used for adductor spasmodic dysphonia in patients with previous recurrent laryngeal nerve section. ⁴⁴ Analysis of treatment results in this population suggests that injection of muscles on the side of the previous nerve section tends to optimize results and minimize adverse effects. ⁴⁵ Ludlow and colleagues ⁴⁶ reported a significant reduction in all speech symptoms using bilateral thyroarytenoid injections in the setting of previous recurrent laryngeal nerve section.

In patients with abductor spasmodic dysphonia, the target muscles for denervation are the PCA muscles. In our practice, these injections are generally not performed simultaneously to minimize airway risk from impaired abduction. The larynx is reexamined 2 weeks postinjection to determine the therapeutic effect and airway patency, and plan the contralateral injection, usually at half the initial dose. Using this technique, Blitzer et al ⁴⁷ noted that patients achieved 70.3% of normal voice. In other centers, bilateral simultaneous injections may be the norm: simultaneous injections have been shown to be reasonably safe by Stong and colleagues, ⁴⁸ who performed a series of simultaneous bilateral PCA injections without airway complications.

The interarytenoid muscle was found to be dysfunctional in certain cases by Hillel and colleagues ⁴⁹ using five-lead EMG. They found that the interarytenoid muscle can be injected in conjunction with the thyroarytenoid muscle to treat certain cases of refractory spasmodic dysphonia.