P. Marco Fisichella, Nathaniel J. Soper, Carlos A. Pellegrini and Marco G. Patti (eds.)Surgical Management of Benign Esophageal Disorders2014The ”Chicago Approach”10.1007/978-1-4471-5484-6_7
© Springer-Verlag London 2014
7. Head and Neck Manifestations of Gastroesophageal Reflux Disease
(1)
Section of Otolaryngology – Head and Neck Surgery, University of Chicago Medical Center, 5841 S. Maryland Ave., MC 1035, Chicago, IL 60637, USA
Abstract
Laryngopharyngeal reflux (LPR) occurs when gastric contents pass through the upper esophageal sphincter (UES) into the upper aerodigestive tract (UADT). These relatively brief episodes can have sinister implications, resulting in irritation to the delicate mucosa of the larynx, pharynx, Eustachian tubes, and nasal passages. The subsequent inflammatory changes that take place are responsible for many of the signs and symptoms that have come to define the disease process. The majority of patients with LPR lack the classic gastroesophageal reflux (GER) symptoms of heartburn, and dysphagia, therefore making the diagnosis more challenging (Koufman, Laryngoscope, 101:1–78, 1991). Instead, patients with LPR commonly present with a constellation of symptoms reflective of UADT inflammation including chronic cough, hoarseness, and postnasal drip (Fennerty, Gastroenterol Clin North Am, 28:861–873, 1990; Koufman, Gastroesophageal reflux and voice disorders. In: Rubin J (ed) Diagnosis and treatment of voice disorders. Igaku-Shoin, New York, pp 161–175, 1995). Unfortunately, the relative ubiquity of these nonspecific symptoms makes the diagnosis of LPR difficult to establish based solely on clinical presentation. In turn, further workup is often necessary, requiring a combination of laryngoscopy (Fennerty, Gastroenterol Clin North Am, 28:861–873, 1990; Maronian et al., Laaryngology, 110:606–612, 2001), dualprobe pH monitoring (Muderris et al., Arch Otolaryngol Head Neck Surg, 135:163–167, 2009), and multichannel intraluminal impedance (MCII) testing (Kawamura et al., Am J Gastroenterol, 99:1000–1010, 2004; Hoppo et al., J Gastrointest Surg, 16:16–25, 2012). While each of these tests provides important diagnostic data, they are not without shortcomings. In fact, an accurate diagnostic tool for LPR has yet to be identified, and clearly defined diagnostic criteria have yet to be agreed upon. While the workup of LPR remains controversial, the treatment is more widely accepted and employs a combination of behavioral changes and medical management including PPIs, H2 blockers, mucosal cryoprotectants, and prokinetic agents. Response to therapy is often variable, and many patients require aggressive antiacid intervention to have complete resolution of symptoms (Vaezi, Nat Clin Pract Gastroenterol Hepatol 2:595–603, 2005). Despite these aggressive approaches, there are subsets of patients deemed refractory to medical management. In these cases, surgical intervention may be needed to achieve complete or even partial resolution of symptoms (Oelschlager et al., Gastrointest Surg, 6:189–194, 2002; Brown et al., Surg Endosc, 25:3852–3858, 2011).
Keywords
Laryngopharyngeal reflux (LPR)Upper esophageal sphincter (UES)LaryngoscopyHoarsenessCoughIntroduction
Laryngopharyngeal reflux (LPR) occurs when gastric contents pass through the upper esophageal sphincter (UES) into the upper aerodigestive tract (UADT). These relatively brief episodes can have sinister implications, resulting in irritation to the delicate mucosa of the larynx, pharynx, Eustachian tubes, and nasal passages. The subsequent inflammatory changes that take place are responsible for many of the signs and symptoms that have come to define the disease process. The majority of patients with LPR lack the classic gastroesophageal reflux (GER) symptoms of heartburn, and dysphagia, therefore making the diagnosis more challenging [1]. Instead, patients with LPR commonly present with a constellation of symptoms reflective of UADT inflammation including chronic cough, hoarseness, and postnasal drip [2, 3]. Unfortunately, the relative ubiquity of these nonspecific symptoms makes the diagnosis of LPR difficult to establish based solely on clinical presentation. In turn, further workup is often necessary, requiring a combination of laryngoscopy [2, 4], dualprobe pH monitoring [5], and multichannel intraluminal impedance (MCII) testing [6, 7]. While each of these tests provides important diagnostic data, they are not without shortcomings. In fact, an accurate diagnostic tool for LPR has yet to be identified, and clearly defined diagnostic criteria have yet to be agreed upon. While the workup of LPR remains controversial, the treatment is more widely accepted and employs a combination of behavioral changes and medical management including PPIs, H2 blockers, mucosal cryoprotectants, and prokinetic agents. Response to therapy is often variable, and many patients require aggressive antiacid intervention to have complete resolution of symptoms [8]. Despite these aggressive approaches, there are subsets of patients deemed refractory to medical management. In these cases, surgical intervention may be needed to achieve complete or even partial resolution of symptoms [9, 10].
Pathophysiology
On a mechanical level, it is a commonly reported and held belief that the UES is a critical site of dysfunction in LPR. This makes sense given that the UES is the last barrier before refluxate enters the pharynx. However, experimental evidence has yet to definitively explain the role of the UES in LPR episodes. Cricopharyngeal (CP) electromyography studies of healthy controls and LPR patients failed to demonstrated abnormal activity in LPR patients [11]. Compared to healthy controls, LPR patients do demonstrate greater duration of CP contraction in response to swallowing, but this may be a result of chronic irritation from acid rather than a cause of pharyngeal acid exposure [11]. Patti and colleagues studied with esophageal manometry and dual-probe pH monitoring 70 patients with symptoms of GER and found lower UES resting pressures in patients with greater acid exposure in the proximal esophagus [12, 13]. In these studies, the relationship of UES pressure to pharyngeal acid exposure was not explored directly.
Oelschlager studied 15 patients with extraesophageal symptoms prior to surgical intervention and found resting UES pressures to be lower in the 9 patients with evidence of pharyngeal reflux based on pH-monitoring readings below 4 in this area [9]. Interestingly, in his series restoration of the LES competence by laparoscopic fundoplication eliminated pharyngeal reflux, despite no interventions at the UES.
Using dual-probe impedance and pH monitoring, Kawamura demonstrated the presence of acidic gas reflux in patients with reflux-associated laryngitis but not in GERD patients or controls [6]. This important finding has shed new light on the mechanisms of LPR and its independence from GERD. Given the at least partly gaseous nature of this exposure and the purported small frequency and duration of reflux necessary to result in LPR symptoms, it may be that baseline abnormalities in the UES are a component, but not the sole driver of LPR pathophysiology.
GER and LPR are both associated with esophageal dysfunction. GER episodes are associated with frequent LES relaxation episodes or less commonly, LES hypotension. Esophageal dysmotility also plays an important role in GER, with studies demonstrating that the greater the degree of esophageal dysmotility, the more severe the GER [14]. Abnormal esophageal peristalsis also plays a role in LPR. When studied by pH monitoring, patients with LPR do have decreased esophageal acid clearance as compared to healthy controls [15]. Furthermore, in a prospective study of esophageal motility in 100 patients with symptoms and findings of LPR, hypertensive LES was present in 23 % of patients, and esophageal motility was abnormal in 71 % [16]. Thus, dysmotility and LES abnormalities are appropriate to explore in LPR patients and represent potential targets for management of the disease.
The primary explanation for the laryngeal manifestations of LPR is that the laryngeal mucosa is more susceptible to injury from acid and activated pepsin than esophageal mucosa. In patients with GER, esophageal mucosa demonstrates increased expression of carbonic anhydrase, an enzyme that acts to neutralize luminal pH [17]. Conversely, in patients with laryngeal disorders related to LPR, 64 % demonstrated absent or decreased carbonic anhydrase expression suggesting a decreased ability of laryngeal mucosa to protect itself against the assault of acid exposure [18]. Furthermore, laryngeal tissue exposed to LPR demonstrated downregulation of E-cadherin, a key component of epithelial barrier function and a checkpoint on the pathway to carcinomatous changes [18]. Changes can occur even with short exposure (<60 s) to weak acid–pepsin solutions, as evidenced by in vitro upregulation of messenger RNA for stress response genes EGR-1 and ATF-3 as well as for genes implicated in neoplastic processes such as VEGF and MMP-2 [19].
Highlighting the multifactorial nature of reflux irritation, nonacidic pepsin has also been shown to be toxic above the UES. Cultured hypopharyngeal epithelial cells subjected to human pepsin at a pH of 7.4 demonstrated upregulation of stress- and toxicity-related genes [18]. To date, similar molecular evidence is not available for other tissues in the UADT, but as discussed in the next section, epidemiologic evidence connects LPR to diseases in these areas.
This heightened sensitivity of acid exposure in laryngeal versus esophageal mucosa underlies the key differences between GER and LPR. First, much fewer episodes of acid exposure are associated with LPR compared to GER. Second, the manifestations of LPR occur in many cases in the absence of signs or symptoms of esophageal disease. A landmark study by Koufman demonstrated that as few as three episodes of laryngeal reflux per week can be sufficient to cause severe laryngeal inflammation and injury [1]. Using 24-h pH monitoring, he demonstrated abnormal reflux in 62 % of 182 patients with laryngitis, dysphagia, chronic cough, stenosis, and laryngeal carcinoma. Of these patients, only 43 % reported symptoms of heartburn or regurgitation, supporting the concept of LPR as a distinct clinical entity.
In addition to direct effects on UADT mucosa, acidification of the esophagus can result in indirect changes to the airway. This is possibly due to a vagal-mediated reflex, but research is still evolving in this area. In one recent study using an anesthetized cat model, instillation of physiologic concentrations of hydrogen chloride (HCl) resulted in decreased diameter of small bronchioles, decreased mucociliary clearance, and increased mucous production [20]. These changes are presumed to be part of a protective mechanism in preparation for possible tracheal acid exposure. Despite minimal effect of esophageal acidification on airway resistance in this cat and also in rabbit models, increased mucous production may alone account for some of the laryngeal and airway symptoms attributed to LPR [20, 21]. A similar reflex in the sinuses has also been demonstrated experimentally in humans using a catheter to infuse the distal esophagus, but the response could not be directly linked to acid exposure, as infusion with saline also resulted in mucous production [22].
Epidemiology and Clinical Manifestations
It has been estimated that as many as 25–40 % of adult Americans experience symptomatic GER at least once per month [23]. In turn, it is not surprising that GER has garnered the majority of the attention from both clinicians, as well as the general public, when it comes to manifestations of reflux disease. LPR has only recently gained attention as an important contributor to reflux-related morbidity, based on evidence of connections to several disease processes. These disease processes are evident in three primary areas of the head and neck: the larynx, the nasal cavity, and the middle ear.
Laryngeal
Given the proximity of the larynx to the UES, it is unsurprising that the larynx is the most studied and most frequently encountered anatomic subsite affected by LPR. A 2002 international survey by the American Bronchoesophagological Association revealed that the most common LPR symptoms reflected laryngeal insult and included throat clearing (98 %), persistent cough (97 %), globus pharyngeus (95 %), and hoarseness (95 %) [24]. Coughing/throat clearing has long been recognized as one of the more common presenting symptoms of LPR. In Koufman’s 1991 landmark study, 87 % of patients with proven LPR reported cough or excessive throat clearing. Interestingly, only 3 % of patients with GERD reported similar symptoms [1]. The association between reflux and chronic cough has been demonstrated within several large population-based surveys [25, 26]. The pathogenic mechanism is primarily believed to be micro aspiration of gastric contents into the airway, although vagal-mediated airway reflexes have also be proposed.
In addition to chronic chough, both hoarseness and dysphonia are also common manifestations of LPR. The symptoms are often grouped together since both reflect underlying laryngitis. Like many laryngeal manifestations of LPR, both conditions are commonly encountered within the general population, and most cases do not warrant further diagnostics. However, when symptoms last longer than 2 or 3 weeks, further workup is warranted. In such cases, it is not uncommon to identify reflux as either an inciting etiologic agent or an underlying comorbidity. In 1991 Koufman found it to be the most common manifestation of LPR, occurring in 92 % of LPR patients surveyed [1]. While subsequent studies have argued that this figure might be inflated, it is established that hoarseness is one of the more common presenting symptoms of LPR [27]. Furthermore, a 2005 Cochrane Review found LPR to be one of the more common causes of hoarseness, with as many as 50 % of hoarse patients having some underlying level of acid reflux [28]. Clinical data support the aforementioned epidemiologic data. Wiener and colleagues reported that 78 % of 32 patients with voice complaints had LPR documented by pH monitoring [28]. While not all patients with LPR exhibit dysphonia, those who do have been found to have more episodes of proximal reflux [29].
While cough and dysphonia remain perhaps the most common laryngeal presentations of LPR, globus sensation is also a common complaint amongst LPR patients. Classically referred to as having a “lump in one’s throat,” globus sensation has been estimated to be present in as many as 63 % of patients with reflux symptoms [30]. Furthermore, studies examining patients with globus sensation have found that nearly 72 % had LPR as diagnosed by either symptom score or pH monitoring [31].
Although rare, paroxysmal laryngospasm can be one of the more frightening and stressful manifestations of reflux disease. Defined as the tonic and sustained contraction of the thyroarytenoid muscle, laryngospasm ultimately stems from the sudden, prolonged, and forceful adduction of the vocal cords. Often described as “throat closing,” patients commonly report an inability to inspire despite adequate and conscious attempts to do so. Reflux-induced laryngospasm was first described by in 1977 [32]. Since its original description, numerous other studies have gone on to support the hypothesis that LPR is a significant etiologic agent in the pathogenesis of spontaneous laryngospasm. Loughlin and Koufman conducted a prospective study of patients with laryngospasm and demonstrated that over 92 % had evidence of reflux-related disease on physical exam and 83 % had abnormal ambulatory pH testing [33]. While extremely distressing for the patient, laryngospasm responds well to antireflux medication. In Loughlin’s study, 100 % of the patients with LPR and laryngospasm responded to 6 weeks of antireflux medication [33].
Laryngeal findings in LPR range from subtle signs of inflammation to more severe manifestations of vocal cord granulomas and laryngeal stenosis. Vocal cord granulomas arise from the posterior cartilaginous vocal process of the true vocal cord and are considered an inflammatory response from injury to the underlying perichondrium. While laryngeal granulomas have been recognized in a variety of patients, those arising from reflux are most classically seen within adult males in the fourth or fifth decade of life [34]. Patients often present with dysphonia, globus sensation, hoarseness, cough, and sore throat. While reflux is thought to be an important instigator in the pathogenesis of vocal cord granuloma, one-to-one correlation has not been proven [35], and the incidence of vocal cord granuloma is relatively low. Thus, the presumptive hypothesis offered by many researchers is that rather than acting as a primary etiologic agent, LPR may act as a priming event or co-instigator in the pathogenesis of vocal cord granuloma.
Laryngeal and subglottic stenosis are also associated with LPR [4]. A single-center study demonstrated that over 50 % of patients with idiopathic subglottic stenosis have pepsin embedded in tissue of larynx or subglottic scar. The aberrant localization of gastric-specific protein has led researchers to question whether LPR may be responsible for underlying airway stenosis. Other studies have strengthened this association. One such prospective study investigated a small cohort of individuals with diagnosed subglottic stenosis. These patients had a variety of underlying disease pathologies presumed to be the causative etiology for the stenosis. However, dual-probe pH monitoring revealed that 86 % of the patients investigated had pH readings below 4 at the level of the laryngeal inlet. Such findings have caused researchers to hypothesize that while LPR may not be the sole instigating event in the pathogenesis of subglottic stenosis, it may very well act as a priming or contributing causative factor.
Of all the aforementioned manifestations of reflux disease, none remains more debated than the association between reflux and upper airway malignancy. For over 50 years, gastric reflux has been a popularly implicated etiologic agent of squamous cell carcinoma within the upper airway [36]. Unfortunately, while the association is pronounced, the causality of association has eluded researchers. Proponents of this theory hypothesize that chronic repetitive chemical stimulation of gastric acid leads to laryngeal mucosal dysplasia and there is molecular evidence to support carcinogenic changes due to acid exposure [18, 19]. However, to date no study has been able to demonstrate a causal relationship between reflux and laryngeal malignancy [37].
Nasal
For the past two decades, researchers have hypothesized that LPR may be an important etiologic agent within the pathogenesis of chronic sinusitis. While an association between the two pathologies exists, studies have failed to consistently establish casualty between the two conditions. Opponents of this theory maintain that both conditions are extremely common, and therefore, they attribute any association to confounding bias.
The association between reflux and chronic rhinosinusitis is primarily supported by retrospective case analyses or prospective cohorts trials without controls – the vast majority lacking significant power and subject to significant biases [38]. Despite these pitfalls, the findings of these studies demonstrate intriguing results. For instance, investigations of children with medically refractory chronic sinusitis found that nearly 63 % of pediatric patients had reflux identified on pH monitoring [39]. A separate study investigated children who were on a waiting list for sinus surgery and revealed that 73 % had LPR identified on pH monitoring, of which 41 % had no other LPR symptomology [39]. Although the data from each study are compelling, both had small sample sizes, 30 patients and 22 patients, respectively. Moreover, each study investigated pharyngeal reflux as monitored by probes placed in the hypopharynx. While the data provided by these studies may be diagnostic of LPR, they do not prove the presence of nasopharyngeal reflux. Instead, the authors seem to assume that LPR and NPR are synonymous.
Adult studies have investigated the presence of NPR in patients with medically refractory chronic rhinosinusitis (CRS). One study found that NPR events (pH < 5) occurred significantly more often in CRS patients when compared to healthy adults (76 % vs 28 %) [40]. Unfortunately, a separate study of CRS patients using 4-channel pH probe found that while 32 % of CRS patients had pH probe evidence of GERD, only 3 % had LPR, and just 0.2 % had evidence of nasopharyngeal reflux [41].
While studies investigating the association between LPR and CRS are flawed, studies investigating the effect of antireflux therapy on CRS symptoms are perhaps even more poorly controlled. Like the aforementioned studies, biases and confounding variables make it difficult to draw conclusions from the data. For instance, Kleeman and colleagues investigated the utility of PPI therapy on CRS patients who continued to have symptoms 3 weeks after functional endoscopic sinus surgery. The study failed to account for postsurgical improvement within the treatment group and had no control group. However, the authors noted that the addition of PPIs reduced nasal symptoms in 76 % of patients. Other studies also relied on patient symptom scores to monitor for nasal symptoms improvement after institution of PPI therapy [40, 42]. While this study showed that 93 % of patients with medically refractory CRS had improvement after 1 month of PPI, it is subject to patient subjectivity and recall bias [42].
Otologic
LPR may also be a contributing factor to otitis media, particularly in children. When compared to adults, the pediatric Eustachian tube is both shorter and more horizontal. This anatomy makes the inner ear susceptible to migrating infections from the nasopharynx, and it is the generally accepted explanation for the increased incidence of otitis media with effusion (OME) within the pediatric population. The same pathogenic model used to describe the migration of microbes from nasopharynx to inner ear has been applied to reflux as well. The relatively close proximity between UES and Eustachian tube in the pediatric patient means that reflux gastric contents can easily reach the nasopharynx, Eustachian tube, and theoretically the middle ear. A number of studies have demonstrated an association between LPR and otitis media. Analysis of pediatric LPR patients found that reflux was present in 12.6–64 % of patients with chronic OME and 61.5–64.3 % of recurrent acute otitis [43–46]. Several elegant studies have gone on to support the underlying epidemiologic data by searching for pepsin and pepsinogen within the middle ear effusion product [47–49]. One such study examined both adenoid tissue and middle ear effusions from patients undergoing tympanoplasty and adenoidectomy. The study demonstrated that pepsinogen was detected in 84 % of patients with middle ear effusions, at concentrations 1.86–12.5 times higher than that of serum. A separate, but similarly designed, study revealed the presence of pepsin and pepsinogen within 59 of 65 middle ear effusion samples [48].
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
