a broader constellation of symptoms, including neurogenic cough, throat clearing, dysphonia, and dysphagia.5 The suggestion that cough could be a manifestation of neuropathy or laryngeal hypersensitivity was first described as a part of ILS by Morrison et al (1999). The authors hypothesized that an acute traumatic event or a chronic noxious stimuli placed the larynx into a hyperfunctional, spasm-ready state.3 Patients with ILS exhibit evidence of laryngeal hyperfunction (laryngospasm, dysphonia, globus, and cough) that can be reliably triggered by innocuous sensory stimuli (odors, food, voice use, exertion, or refluxate).3 Although the most common proposed etiologic factor was gastroesophageal reflux, 17 out of 39 patients had a preceding viral illness.3 Subsequent investigations have focused on the neurogenic manifestations of these laryngeal symptoms and evidence of vagal neuropathy. Linking the concepts of a neuropathy and antecedent viral illness, case series have described a postviral vagal neuropathy (PVVN) with symptoms that may include globus sensation, neurogenic chronic cough, vocal fold paresis, and dysphagia.6–8 One of these studies, by Amin and Kaufman (2001), demonstrated laryngeal electromyographical evidence of neuropathy. Descriptions of PVVN have inclusion criteria of an antecedent viral illness, but other descriptions of ILS and laryngeal sensory neuropathies exhibit cases of vagal neuropathy without preceding viral illness.3, 6, 7, 9, 10 In addition, Giliberto et al (2017) looked at predictors of response to treatment for neurogenic cough and noted no association between preceding viral illness and improved response to neuromodulator therapy, although the study population was small.10
Neurogenic cough associated with a PVVN limits the discussion of neurogenic coughs to those suspected to be viral in origin. In contrast, laryngeal sensory neuropathy (LSN) is a more general term used to describe chronic cough associated with evidence of vagal neuropathy.1, 9, 11–13 Unfortunately, beyond subjective patient reports of laryngeal dysesthesias (“tickle”), objective evidence of vagal sensory neuropathy is limited or not readily available. Some authors have used capsaicin sensitivity testing, but changes in cough threshold are noted in the chronic cough population and not specific to patients with neurogenic cough.14–16 Furthermore, limiting the discussion of cough as a sensory neuropathy fails to acknowledge that motor neuropathy was noted in 49% to 100% of patients in patient with both LSN and PVVN.6, 7, 9 These studies have improved our current understanding of the interplay of neuropathy and chronic cough. With varied terminology in the literature, most authors favor the inclusive terminology of neurogenic cough for the symptoms of chronic cough thought to be related to vagal neuropathy and laryngeal hypersensitivity syndrome to encompass the syndrome of hyperfunctional laryngeal symptoms including cough, globus, throat clearing, and muscle tension associated with innocuous stimuli.4, 17–19
The act of coughing can be both a reflex requiring little conscious control and a voluntary act.5 A discussion of neurogenic cough benefits from a brief review of the cough reflex and how it is thought to be affected in laryngeal hypersensitivity syndromes. The reflex is initiated by excitation of the sensory peripheral branches of the vagus nerve, originating from the mucosa of the larynx, pharynx, and esophagus.20 Both noxious chemical stimuli (capsaicin or odors) and pressure can activate the afferent sensory nerve via chemosensitive nociceptors and mechanosensors, respectively.5 In the healthy individual, these excitatory stimuli travel up afferent first-order neurons to the solitary nucleus in the brainstem.4, 5 Little is known about the exact physiology of sensory input integration and summation in the brainstem. Evidence in animal models demonstrates that frequent sensory nerve action potentials recruit N-methyl-D-aspartate (NMDA) receptors expressed on second-order nerves in the brainstem.21 These second-order neurons are responsible for coordination of the respiratory motor response (ie, cough), but it is unclear how this may be augmented in the patient with neurogenic cough. Connections from the motor and sensory cortex to the solitary nucleus allow for volitional initiation and inhibition of coughing.5
It is hypothesized that in patients with neurogenic cough, this normal cough reflex is potentiated by three mechanisms. The first proposed mechanism involves lowering the threshold for activation of afferent sensory nerves. Antigen or viral inflammation in animal models leads to increase neuropeptide production and expression, thereby influencing the concentration of mechanoreceptors and nociceptors.5 These changes can lower the threshold for activation of the sensory afferent nerve. In addition, these increased receptor expressions can lead to spontaneous activation, which underlies the second proposed mechanism. The sensation of phantom irritation of the throat, or the “bogus tickle,” as stated by Bastian et al,11 is often clinically noted in the setting of neurogenic cough. In the chronic pain literature, changes in afferent nerves, as noted above, can lead to spontaneous activation. However, similar spontaneous depolarizations have not been demonstrated in a cough model, nor have they been linked to laryngeal dysethesias.5 A final mechanism of neurogenic cough involves an increase in the gain of the cough reflex at the brainstem level (or central sensitization).4, 5 Convergence of airway and esophageal sensory inputs can amplify incoming afferent sensations, augmenting the gain of the reflex. In addition, neuropeptides, such as substance P, have been shown to be upregulated following neural inflammation, and can produce durable excitatory effects on afferent nerves.5 These proposed mechanisms exclude lesser-studied effects from muscular spasm and irritation from overstimulation, and do not include behavioral components such as habit, neurosis, and secondary gain.22
As neurogenic cough has often been a diagnosis of exclusion, the focus of initial history and physical exam should be on symptoms and signs of other common etiologies, as detailed in previous chapters. It is important to keep in mind that many of the symptoms of a vagal neuropathy can be nonspecific and can cloud the clinical picture. For example, vagal nerve dysfunction can lead to dysphagia and esophageal dysmotility, resulting in heartburn and solid food dysphagia—symptoms more often associated with gastroesophageal reflux disease (GERD) than neurogenic cough. Reviewing the literature, prospective cohorts have identified elements of the patient history and physical exam that are associated with neurogenic cough. These elements include reliable cough triggers, antecedent viral illness, and evidence of laryngeal hyperfunction on physical exam. Although these elements may have limited sensitivity and specificity for neurogenic cough, they have evidence to support their association.
The description of hypersensitive larynx syndrome and ILS both cite benign triggers for neurogenic cough. These triggers include often innocuous stimuli such as odors, food, exercise, temperature changes, or speech.3–5 As discussed in the pathophysiology section, a lower threshold for activation of afferent nerves of the larynx and pharynx could explain an overzealous reaction of the cough reflex to typically innocuous stimuli. Along with cough response to innocuous stimuli, the description of laryngeal hypersensitivity syndromes is associated with symptoms of globus, dysphonia, and laryngospasm. In Morrison’s landmark paper describing ILS, 65% (25 out of 39) of patients had laryngospasm, 26% (10 out of 39) of patients experienced dysphonia, and 77% (30 out of 39) patients experienced cough or globus. Although present, these symptoms are not specific enough in the setting of chronic cough to suggest a neurogenic origin.
Patients with PVVN by definition have an antecedent vial illness, but other syndromes such as ILS cite antecedent viral illness in 17 out of 39 cases.3, 6–8 Many experts feel that acute onset of cough or other nonspecific symptoms of ILS following an upper respiratory illness can support a neurogenic origin of chronic cough. The larger PVVN description by Rees et al noted patients had persistent cough (52%), throat clearing (48%), dysphonia (41%), and other less-common complaints after upper respiratory illness.7 Given that the presence of a preceding viral illness is required for inclusion in these studies, we cannot make conclusions about the frequency or association of neurogenic cough with antecedent upper respiratory illnesses. Other case series, not limited to PVVN, inconsistently report antecedent URI or do not have sufficient patient populations to note an association. A small retrospective study of patients suspected to have neurogenic cough looked at factors associated with response to neuromodulator therapy and did not demonstrate an association between viral illness and response to therapy; however, few patients in this study were reported to have an antecedent URI (four out of 25).10
Similar to the history, the absence of objective findings on physical exam of other etiologies of chronic cough is most consistent with neurogenic cough. Clinicians trained and familiar with palpation of the laryngeal muscles may find tension in the strap muscles, especially in the thyrohyoid space, suggesting laryngeal hyperfunction consistent with the description of, although not specific to, ILS.3 Evidence of other upper cranial neuropathies may support the presence of a vagal neuropathy.6 Evidence of motor neuropathy of the larynx and pharynx can be quite informative and will be discussed in the next section. Otherwise, there are few specific physical exam findings consistent with neurogenic cough.
In summary, there are myriad of clinical symptoms associated with neurogenic cough and laryngeal hypersensitivity syndrome. Complaints of globus, dysphonia, dysphagia, and throat clearing are not, in isolation, specific enough to indicate a neurogenic etiology of chronic coughing. However, there are historical elements and symptoms that, when taken in context of the whole clinical picture, may support the diagnosis of neurogenic cough. Expert reviews cite the presence of triggers (chemical, odors, mechanical, or temperature)4 or antecedent viral illness3, 6, 7 as the most specific elements of the history and physical supporting a diagnosis of neurogenic cough.
Successful diagnosis of neurogenic cough requires laboratory and diagnostic testing to appropriately assess the wide differential of chronic cough. As with patient history and physical exam, negative findings for other etiologies of chronic cough can support a diagnosis of neurogenic cough. Chapter 9 will detail a comprehensive diagnostic approach to chronic cough, including the use of objective testing to evaluate pulmonary, sinonasal, and reflux-related causes in the context of the individual patient. Particularly germane to neurogenic cough is an evaluation of the larynx to rule out pathology and to evaluate for evidence of motor neuropathy. This can be accomplished with laryngoscopy with or without stroboscopy and LEMG. For providers not familiar with these procedures, the next section will give a brief overview of each procedure before moving on to how they can help with diagnosis and treatment.
LARYNGOSCOPY AND STROBOSCOPY
Visualization of the larynx can be performed indirectly with a mirror, or directly visualized through the use of a flexible fiberoptic laryngoscope or a rigid 70-degree endoscope. Although a trained and experienced examiner may be able to obtain a crisp and clear view of the larynx with a mirror through indirect laryngoscopy, the ability to record and obtain reliable images with superior image resolution has led many otolaryngologists to favor either flexible fiberoptic, distal chip digital, or rigid transoral laryngoscopy.23 Flexible laryngoscopy is the more common modality and is well tolerated by patients in the setting of an office visit, with or without topical anesthesia to the nose.23, 24 Once the scope traverses the nose, it provides a view of the oropharynx, hypopharynx, and larynx. The scope delivers light via fiberoptic fibers, and the images are returned by additional fibers or a distal CCD chip via wire. These images can be recorded on video capture systems, allowing for slow-motion review and storage for later comparison. Rigid Hopkins rod laryngoscopes (70- or 90-degree) are introduced through the mouth over the tongue and visualize the larynx with a camera attached as an image-capture device. Laryngoscopy (either flexible or rigid) can be performed using a special lighting protocol called stroboscopy. In this case, light from the scope is pulsed at a frequency just slightly different than the patient’s fundamental frequency. This pulsed light provides a sequence of images that can give the illusion (or simulation) of a slow-motion view of the glottic vibratory cycle as long as the patient can maintain a consistent and periodic vibration.23 The addition of this simulated slow motion allows for assessment of vibration of the vocal folds and their closure pattern. The improved optics allow for assessment of mucosal lesions and subtle differences in motion. Although rigid endoscopes provide excellent video quality of the vibrating vocal folds during stroboscopic light imaging, they cannot assess vocal fold motion during connected speech.23, 24
Findings on laryngoscopy and stroboscopy are subjective; however, certain diagnoses have different interrater reliabilities. While many examiners can agree on the side of vocal fold immobility and paralysis, evidence of subtle paresis is more difficult to ascertain as to which is the affected side.25, 26 Signs on stroboscopy of vocal fold paresis include asymmetry in mucosal wave, motion asymmetry, laryngeal axis rotation, compensatory supraglottic hyperfunction, and glottic insufficiency.25, 27 Although motion or vibratory abnormalities are not reliable predictors of side of injury, both Simpson et al (2011) and Woo et al (2016) showed that physicians correctly diagnosed the presence of asymmetry stroboscopy 83% and 88% of the time, respectively.25, 26 Thus, when performed by an experienced provider, laryngoscopy with stroboscopy has a good association with the presence of neuropathy, but poor reliability as to which side is weak.
Many cough patients benefit from the involvement by an otolaryngologist or laryngologist for evaluation of the larynx and hypopharynx. A flexible exam with stroboscopy by an experienced interpreter will likely provide a thorough assessment of vocal fold motion asymmetries consistent with motor neuropathy. The presence of motor neuropathy on stroboscopy or, as will be discussed in the next section, on LEMG can support the presence of a vagal neuropathy. This may be helpful in narrowing the differential diagnosis and supporting a diagnosis of neurogenic cough.
LEMG is a procedural evaluation technique of muscle fiber action potentials and motor unit action potentials in the laryngeal muscles.28 Typically, it is performed under local anesthesia in the office by a laryngologist, neurologist, or physical medicine and rehabilitation electromyographer, or two of these providers together. After local anesthesia, a needle electrode is advanced into the laryngeal muscles and morphology of the action potentials are qualitatively or quantitatively evaluated for signs of acute or chronic injury,28, 29 the details of which are beyond the scope of this chapter. Some groups perform qualitative analysis to help quantify degree of injury and standardize diagnosis. However, the practice of LEMG remains varied as does the ability of providers to incorporate it into a regular outpatient clinic, as some providers may need to schedule an LEMG weeks or months following an initial evaluation. Although the qualitative assessment of motor neuropathy on LEMG has a subjective component, the data from LEMG are felt to be more objective than the finding of laryngoscopy with stroboscopy. LEMG is often used as the gold standard to compare laryngoscopy findings when evaluating sensitivity and specificity of clinician ratings for subtle paresis.25, 26
Motor Neuropathy and Choosing the Diagnostic Modality
Stroboscopy and LEMG can be used to assess a patient with neurogenic cough for associated motor neuropathy that may suggest a neurologic change has occured. Studies of neurogenic cough often include videostroboscopy results,10, 13, 30 but others include evidence from both LEMG and stroboscopic modalities.6, 9 In the two retrospective case series and one prospective consecutive cohort that looked at response to medical treatment, all studies demonstrated improved response to treatment in patients with motor neuropathy (Table 6–1). In considering the decision for LEMG and stroboscopy, the latter is far more readily available than LEMG in the evaluation of neurogenic cough. In addition, beyond looking for the presence of motor neuropathy, flexible laryngoscopy and stroboscopy can be helpful in focusing the differential and ruling out laryngeal pathology as the source of the chronic cough. Thus, given its availability, tolerability, and benefit to the cough differential, stroboscopy is often performed in the evaluation of chronic cough.
Table 6–1. Motor Neuropathy and Neuromodulator Response. Review of the studies that evaluated patients for evidence of motor neuropathy and response to neuromodulator therapy
LEMG = Laryngeal electromyography.
*Proportion difference reported without statistical evaluation.