Reflux is a significant contributor to cough in otolaryngology practice; cough is just one marker of its many negative effects on the upper aerodigestive tract. Reflux causes cough both by direct irritation/inflammation and by increasing sensitivities to other noxious agents. Detailed and diligent clinical evaluation, including laryngoscopy, is useful in advancing the working diagnosis of reflux-associated cough. Supplemental testing, including impedance monitoring of esophageal refluxate, can be important to evaluate for both acidic and nonacidic reflux exposure. The mainstay of treatment continues to be dietary and other lifestyle interventions and drug therapy. Although proton-pump inhibitor therapy is effective in most patients, especially those with acid reflux disease, prokinetic therapy is probably very important with those with combined acid and nonacid disease and those with pure nonacid disease. It is likely that failure to improve can be due to behavioral and drug compliance issues. Antireflux surgery can yield long-lasting positive outcomes in carefully selected patients despite the lower efficacy of treatment for primary upper aerodigestive tract symptoms (cough, hoarseness, sore throat) compared with heartburn and regurgitation.
Laryngopharyngeal reflux (LPR) and gastroesophageal reflux disease (GERD) continue to be important clinical entities in otolaryngology; reflux is one of the most common causes of cough found in referral clinics. This content focuses on the relationship between reflux and cough and expands upon the discussions in other articles in this publication, such as “Cough: a Worldwide Problem” and “Unexplained Cough.” It is also necessary to consider LPR and GERD when discussing other entities in this issue, because patients with troublesome cough may have reflux and rhinitis, for example, perhaps caused by underlying autonomic dysfunction. Albeit somewhat artificial to completely separate them, the following sections will deal only with the entities of LPR and GERD and how cough may manifest as part of these disorders, how they can be identified, and how they may be treated.
Both LPR and GERD are quite common; it is estimated that 7% to 10% of Americans have reflux symptoms every day. Twenty percent suffer at least once a week, and up to 36% suffer once a month. These include many familiar complaints such as heartburn, dysphagia, and dyspepsia, as well as nonspecific complaints commonly seen in otolaryngology practice, such as globus sensation, hoarseness, throat clearing, postnasal drip, and of course, cough, the topic of this article.
Within the otolaryngology community, the evidence clearly reflects an increasing awareness in the diagnosis and treatment of reflux disorders. The direct cost of antireflux medical treatment (predominantly in the form of proton pump inhibitors, or PPIs) is staggering, estimated at over $14 billion annually. This cost estimate does not include the medical cost related to the clinic evaluation and diagnostic testing for this disorder, nor does it include any estimate of indirect costs related to loss of work productivity and effects on quality of life. There is also the possible cost of complications related to the use of PPIs such as osteoporosis and bacterial pneumonitis.
In this article, the pathophysiology of reflux and, more specifically, reflux-associated cough, are discussed. Laryngeal inflammation and irritation, reflex cough, and concepts such as nonacidic reflux are discussed. Treatment options, including pharmacologic therapy and behavioral and speech–language interventions as well as surgery are presented.
Pathophysiology of reflux
At one point or another, most adults have had some complaint of heartburn; the pathophysiology of gastric contents flowing backward into the esophagus and even into the larynx and pharynx above is quite familiar in this sense. Ultimately, it is the failure of the lower esophageal sphincter (LES) to control this unwanted retrograde flow that is at the heart of gastroesophageal reflux. The digestive tract must maintain some capacity for retrograde flow, as it is depended upon for the urgent expulsion of gas or emesis when noxious elements are ingested or there is too much gas in the system. The LES cannot just be an unchangeable, static, one-way valve. Perhaps as a byproduct of this, pathologic reflux can and does occur. There are numerous potential mechanisms, however, by which reflux, GERD, and LPR may be related to cough, as listed in Table 1 .
| Action | Mechanism |
|---|---|
| Regurgitation of gastric contents to the laryngopharynx | Mechanical stimulation |
| pH-sensitive stimulation | |
| Non-acidic reflux precipitating tissue change | |
| Enhanced cough sensitivity to other irritants | |
| Gross and “micro” (or silent) aspiration | Aspiration pneumonia |
| Chemical tracheitis and/or pneumonitis | |
| Distal esophageal reflux | Vagal-mediated reflex |
| Esophago-bronchial reflex | Vagal-mediated |
| Likely cough secondary to pulmonary exacerbation | |
| Cough-induced reflux | Diaphragmatic/LES discoordination or relaxation |
The LES and transient LES relaxation
Fyke and colleagues in 1956 often is credited with an early description and definition of the LES and its function. LES activity reflects a combination of dynamic and more static sphincteric characteristics. The latter include the fibers of the crural diaphragm that create the external sphincter. The more familiar dynamic component comes from the actual fibers of the muscular layer of the esophagus; together these two create the LES zone that is approximately 4 cm in length.
Transient lower esophageal pressure relaxation (TLESR) is believed to be the key abnormality that leads to the pathologic reflux of gastric material (be it from the stomach or duodenum) into the esophagus. The LES is sensitive to intrathoracic and intra-abdominal pressure; as the crural fibers of the diaphragm contract on inspiration, the pressure in the external aspect of the sphincter rises. Cough itself, with its often violent changes in thoracic pressure, also can negatively impact the LES in this manner. In the normal state, the LES varies with inspiration, rising with diaphragmatic contraction due to the extrinsic force of the crural fibers. As an example of the inter-relatedness of the foregut and its derivatives, esophageal stimulation by distension will lead to a drop in LES pressure by the inhibition of the crural diaphragmatic fibers. It is possible that this reflex contributes to the pathophysiology of reflux by way of a positive feedback loop ; stimulation of the distal esophageal afferents in response to irritation or distension causes further decline in LES pressure and, perhaps, more reflux.
The dictum “no acid, no ulcer” that reflected the clinical mindset during the decades of surgical dominance of peptic disease probably is no longer accurate. With the advent of potent and effective medical therapy for acid suppression and advances in diagnostic technology, it now is believed that nonacidic reflux likely also plays a role in the pathogenesis of cough in some refractory patients. The presence of noxious nonacidic agents continues to be offered as an explanation of the PPI treatment failure in many cases. Indeed, there is growing evidence that the entity of nonacidic reflux is quite real and not uncommon; beyond that, its significance continues to be worked out.
The two principal nonacidic agents discussed here are bile and pepsin. Bile has been suspected as a source of laryngeal inflammation and irritation for decades. Some of the reluctance to accept bile as a notable contributor to reflux pathophysiology in the upper aerodigestive tract is because of some mixed messages from initial animal studies. In a superb study, Vaezi demonstrated that bile indeed can injure the laryngeal epithelium but only in an acidic environment. In 2005, Sasaki was able to demonstrate histologic laryngeal injury in a rat model following bile exposure in a neutral environment. Work in this area is ongoing.
The literature regarding pepsin, the principal proteolytic enzyme of the stomach, continues to grow. Although pepsin clearly is predominantly active in acidic pH and has been shown to cause laryngeal injury in this mileu, pepsin retains its proteolytic activity up to a pH of 7, and can be reactivated. Johnston found intraepithelial pepsin in the larynx of patients with the clinical diagnosis of LPR but not in controls ( Fig. 1 ); in these same patients, pepsin was absent in their esophageal epithelium. Furthermore, these patients had a depletion of intracellular carbonic anhydrase (type 3). Pepsin is believed to be taken up into the laryngeal epithelium by an active, receptor-based mechanism. Given that the focus of chemotherapy and even surgery for reflux disease over the past century has focused on the reduction of acid production and neutralization, the possibility that a nonacidic injurious agent may contribute to ongoing damage and symptomatology is a major paradigm shift. Individual variability in patients’ ability to withstand or repair injury also may play a role in understanding LPR, GERD, and specific clinical situations related to these disorders, such as reflux-associated cough.
The LES and transient LES relaxation
Fyke and colleagues in 1956 often is credited with an early description and definition of the LES and its function. LES activity reflects a combination of dynamic and more static sphincteric characteristics. The latter include the fibers of the crural diaphragm that create the external sphincter. The more familiar dynamic component comes from the actual fibers of the muscular layer of the esophagus; together these two create the LES zone that is approximately 4 cm in length.
Transient lower esophageal pressure relaxation (TLESR) is believed to be the key abnormality that leads to the pathologic reflux of gastric material (be it from the stomach or duodenum) into the esophagus. The LES is sensitive to intrathoracic and intra-abdominal pressure; as the crural fibers of the diaphragm contract on inspiration, the pressure in the external aspect of the sphincter rises. Cough itself, with its often violent changes in thoracic pressure, also can negatively impact the LES in this manner. In the normal state, the LES varies with inspiration, rising with diaphragmatic contraction due to the extrinsic force of the crural fibers. As an example of the inter-relatedness of the foregut and its derivatives, esophageal stimulation by distension will lead to a drop in LES pressure by the inhibition of the crural diaphragmatic fibers. It is possible that this reflex contributes to the pathophysiology of reflux by way of a positive feedback loop ; stimulation of the distal esophageal afferents in response to irritation or distension causes further decline in LES pressure and, perhaps, more reflux.
The dictum “no acid, no ulcer” that reflected the clinical mindset during the decades of surgical dominance of peptic disease probably is no longer accurate. With the advent of potent and effective medical therapy for acid suppression and advances in diagnostic technology, it now is believed that nonacidic reflux likely also plays a role in the pathogenesis of cough in some refractory patients. The presence of noxious nonacidic agents continues to be offered as an explanation of the PPI treatment failure in many cases. Indeed, there is growing evidence that the entity of nonacidic reflux is quite real and not uncommon; beyond that, its significance continues to be worked out.
The two principal nonacidic agents discussed here are bile and pepsin. Bile has been suspected as a source of laryngeal inflammation and irritation for decades. Some of the reluctance to accept bile as a notable contributor to reflux pathophysiology in the upper aerodigestive tract is because of some mixed messages from initial animal studies. In a superb study, Vaezi demonstrated that bile indeed can injure the laryngeal epithelium but only in an acidic environment. In 2005, Sasaki was able to demonstrate histologic laryngeal injury in a rat model following bile exposure in a neutral environment. Work in this area is ongoing.
The literature regarding pepsin, the principal proteolytic enzyme of the stomach, continues to grow. Although pepsin clearly is predominantly active in acidic pH and has been shown to cause laryngeal injury in this mileu, pepsin retains its proteolytic activity up to a pH of 7, and can be reactivated. Johnston found intraepithelial pepsin in the larynx of patients with the clinical diagnosis of LPR but not in controls ( Fig. 1 ); in these same patients, pepsin was absent in their esophageal epithelium. Furthermore, these patients had a depletion of intracellular carbonic anhydrase (type 3). Pepsin is believed to be taken up into the laryngeal epithelium by an active, receptor-based mechanism. Given that the focus of chemotherapy and even surgery for reflux disease over the past century has focused on the reduction of acid production and neutralization, the possibility that a nonacidic injurious agent may contribute to ongoing damage and symptomatology is a major paradigm shift. Individual variability in patients’ ability to withstand or repair injury also may play a role in understanding LPR, GERD, and specific clinical situations related to these disorders, such as reflux-associated cough.
How does reflux affect the larynx and upper airway?
The impact of reflux may occur by acute or chronic injury, and by noncontact effects (ie, reflex activation of laryngeal and lower airway responses leading to hoarseness or cough are two examples). Fundamentally, the basic event in reflux-associated cough is believed to be pathologic reflux rising up the esophagus, above the upper esophageal sphincter (UES), resulting in irritation of the larynx. Nonacidic reflux also may act in this manner, as demonstrated by Patterson and colleagues in a study of 37 cough patients, roughly half with asthma and half with cough without asthma. Each subject underwent esophageal impedance and pH probe testing off of acid-suppressive medication. The patients, in general, had some evidence of cough temporally related to reflux events (7 of 26 subjects). Although this fraction is relatively modest, this higher symptom-association group had a higher number and proportion of reflux episodes crossing the UES into the pharynx; there was no difference in the reflux event/symptom association group with regard to esophageal reflux events.
There is no doubt that reflux is a major contributing cause of other physical changes in the larynx, such as vocal fold nodules, granulomata, and pseudosulcus. The more common, albeit less-specific findings of vocal fold edema and erythema are discussed in the section on clinical evaluation.
These physical changes in the larynx associated with reflux are mentioned here to demonstrate the macroscopic impact of pathologic reflux on the upper aerodigestive tract. In Koufman’s landmark 1991 Triological thesis, the disorder that had the highest association with reflux was airway stenosis; many of these patients have chronic cough as part of their clinical presentation. It is not known if the cough often associated with airway stenosis is related to the underlying reflux or by some impact on the lower airway, mucus retention, or generalized inflammation. There is also a strong association between idiopathic pulmonary fibrosis and reflux, as another example of an upper airway disorder whose pathophysiology may result in end-organ changes. Beyond these gross physical changes at the larynx and upper airway, much of the derangement related to reflux and cough occurs at a more microscopic level.
Cough: reflux or reflex?
The mere presence of acid in the esophagus also may stimulate the glottis by way of the esophago-glottal reflex, or perhaps the more central airways themselves. Lang and the group at the Medical College of Wisconsin have reported that, based on their animal model, the main impact of acid reflux on the lower airway was to increase mucus secretion. Furthermore, they postulated that the increase in mucus may act as a buffering agent against refluxed materials in the upper or lower airway.
Javorkova and colleagues from Slovakia published a compelling study in 2008 that further elucidated the relationship between reflux and cough. Twenty-five patients with GERD were recruited and submitted to catheter infusions of acid or saline in a randomized, double-blind manner. Their cough response to inhaled capsaicin was measured in all test conditions. Interestingly, in the nine subjects who had pre-existing GERD with cough, sensitivity to experimental capsaicin exposure was heightened with esophageal acidification. The 15 GERD subjects without cough were not affected by esophageal acidification. Furthermore, this experiment also was run on 18 healthy volunteers; again, no increase in cough sensitivity to capsaicin occurred with esophageal acidification. This indicates that the presence of distal esophageal acidification can increase the reflux cough sensitivity to other, nonacidic and even exogenous agents. It is possible that this is a significant part of the pathophysiology of the irritable larynx syndrome and certainly helps in understanding and managing cough patients.
Ferrari and colleagues in Verona furthered the understanding of the mechanism of reflux-associated cough with a compelling experiment in asthmatic patients. In their work, 29 subjects with known asthma were challenged with methacholine and capsaicin before and after a short course of omeprazole (20 mg twice daily). Of the 29 patients, 17 had acidic reflux events in the distal esophagus; many but not all of the same were in the group of 17 who had proximal reflux events. Omeprazole treatment did not affect these subjects’ reactivity to methacholine challenge, but the response to capsaicin was significantly diminished, but only in patients with pathologic reflux, predominantly in the proximal acid group. The authors concluded that, in asthmatics, omeprazole reduced cough sensitivity in those with proximal reflux without influencing bronchial responsiveness. Another example of the complex interrelation between cough and reflux is seen in a Polish study published in 2005. Ziora and colleagues examined the citric acid cough threshold (CACT) in GERD patients before and after fundoplication. In the preoperative group, CACT was lower than seen in healthy volunteers, indicating that the reflux state increased the sensitivity to outside stimulants beyond whatever direct irritation is provided by the refluxate itself.
Cough in reflux: clinical evaluation
As the understanding of reflux-associated disease of the upper aerodigestive tract evolved, the distinctions between patients with LPR and GERD became clearer. Perhaps the first and most critical point is that the absence of heartburn, the cardinal symptom of GERD, is present in less than half of patients with LPR. An alarming example of the significance of atypical reflux symptoms came from Reavis and colleagues, who reported that cough itself was the symptom most highly associated with the development of esophageal adenocarcinoma, greater than heartburn or other more familiar typical reflux symptoms.
Patients with LPR are not the same as those with GERD; LPR patients tend to have daytime, upright reflux, in contrast to those with the more familiar GERD. On average, the habitus of patients with GERD reveals a higher body mass index (BMI) than normals, which is also higher than those of patients with LPR. Halum and colleagues demonstrated this in a large retrospective study in which patients with LPR had a relatively normal BMI compared with the BMI of subjects with GERD on esophageal pH probe testing in the same clinic.
With regard to the clinical history of patients with cough, Everett and Morice made several potentially interesting observations in an uncontrolled 2007 publication. Within their group of 47 patients with confirmed esophageal reflux and cough, the most common cough characteristics were cough on phonation, cough in the morning when getting up out of bed, and cough around the time of eating. Throat clearing, hoarseness, dysphagia, and globus also were related but less so. Heartburn, so often noted to be present in a minority of patients with LPR, was present in 63% of their patients with reflux and cough. This study was limited, however, in that the authors did not assess if the same findings were present in patients with cough caused by other diseases. Cough is also one of the items on Belafsky’s Reflux Symptom Index, reflecting its clinical weight in making the provisional or working diagnosis of LPR ( Table 2 ).
