BIOMARKERS IN REFLUX AND CHRONIC COUGH
LPR is known as one of several potential etiologies of chronic cough. However, there are limited and mixed data to support both the reliability and specificity of current tools designed to assist in diagnosis of chronic cough and LPR, which range from symptom inventories to laryngoscopic grading schemes.1–4 As such, there has been considerable investigation into the utility of various biomarkers as predictors of LPR-associated chronic cough. In particular, pepsin is produced exclusively by the chief cells of the stomach, and hence its presence outside of the stomach has been posited as a sensitive biomarker for reflux and aspiration of gastric contents.5 In intubated pediatric patients, the sensitivity and specificity of pepsin in bronchoalveolar lavage (BAL) samples for predicting clinically identified aspiration have been found to be up to 80% and up to 100%, respectively.6 Moreover, pepsin has been detected in significantly elevated levels in BAL and tracheal lavage samples acquired from patients with proximal esophageal reflux and chronic cough symptoms, suggesting that BAL pepsin may be a useful measure in differentiating chronic cough secondary to esophageal reflux as opposed to other etiologies.6, 7
Previous literature has supported pepsin in induced sputum as a potential tool for identifying the presence of proximal refluxate (75% sensitive, 91% specific), which may enter the laryngopharynx and contribute to laryngeal hyperresponsiveness and pulmonary disease.8 Given the invasiveness of BAL collection, this represents a compelling potential target, and indeed, several studies, discussed below, have compared the utility of pepsin in salivary and induced sputum samples in the detection of LPR to previously validated methods such as the reflux symptom index (RSI, questionnaire useful in determining the severity of reflux-attributed upper respiratory symptoms),4 reflux finding score (RFS, a tool for grading of laryngoscopic findings indicative of reflux-attributed inflammation),3 and multichannel intraluminal impedance-pH (MII-pH) monitoring, with impedance spanning the esophagus and hypopharynx widely considered ideal for proximal reflux monitoring.
In patients with chronic upper respiratory symptoms indicative of laryngeal hyperresponsiveness—such as chronic cough, globus sensation, dyspnea, and episodic choking—Spyridoulias et al found a specificity of 0.78 for the presence of salivary pepsin as a predictor for inflammatory changes on laryngoscopy, but a sensitivity of only 0.53.9 Notably, pepsin did not significantly correlate with reflux symptom inventory scores or reflux events on MII-pH monitoring, and many patients had discordant results between various testing modalities.9 However, almost half of patients with pulmonary symptoms but minimal inflammatory changes on laryngoscopy had detectable salivary pepsin, perhaps suggesting that mild LPR may be sufficient to cause symptoms of laryngopharyngeal hyperresponsiveness without inducing any visible changes in the airway.9
The utility of pepsin as a biomarker for LPR in adult populations with chronic cough without history of symptoms concerning for gastroesophageal reflux appears less promising. In an unselected group of chronic cough patients, induced sputum pepsin has been found to be inversely related to cough frequency, and furthermore these patients do not have significant amounts of proximal reflux by MII-pH.10 This is suggestive of the notion that in contrast to cough in the setting of LPR, chronic cough secondary to other etiologies may in fact be protective in facilitating the clearance of refluxate and hence preventing microaspiration. These results are consistent with prior studies on pepsin in tracheal lavage specimens, which have found tracheal lavage pepsin to have poor predictive value in an unselected population of patients with chronic respiratory symptoms, but much more sensitive and specific when specifically restricting its use to patients with the aforementioned classic symptoms. Moreover, this may be interpreted as suggestive that LPR is not a major mediator in the pathology of a patient with chronic cough who does not also manifest classic symptoms of GERD, and that in the absence of classical GERD symptoms, evaluation for biochemical evidence of LPR is likely to have limited utility. In stark contrast, others have argued that LPR is in fact the primary etiology of most chronic cough via induction of hyperresponsiveness of the upper airway to noxious stimulation,11 which, if true, would imply that current methods used for detection of pepsin simply do not accurately reflect the underlying reflux events. Additionally, recent studies using hypopharyngeal-esophageal MII with dual pH (HEMII-pH) to diagnose LPR in patients with chronic cough demonstrated that carefully selected patients have a complete or partial response to Nissen fundoplication despite lack of symptom correlation to cough.12 This further supports the theory that nonacid LPR may be a chronic underlying inflammatory condition secondary to pepsin deposition in the laryngopharynx and/or esophagus, on top of which other factors trigger the cough, not necessarily a coordinated reflux event.
The lipid-laden alveolar macrophage index (LLMI), similarly, has been suggested as a marker for aspiration of gastric refluxate as it reflects endocytosis of refluxed food lipids by macrophages in bronchoalveolar lavage specimens. Staining of cells within the BAL fluid with oil red O reveals the lipids that can be quantified with the LLMI, which, in prior studies, has been demonstrated to be up to 100% sensitive but only 57% specific for pulmonary aspiration.13 However, this measure likely has poor utility for investigating chronic cough, as LLMI is not significantly elevated in children with GERD or chronic cough.6 It has been suggested that the LLMI may reflect endocytosis not only of food lipids, but also of lipids from degradation of alveolar phospholipids,6 which may account for its poor specificity. As such, although the LLMI may have some limited value in evaluation of aspiration related disease, its utility in the setting of chronic cough is likely very low.
Numerous challenges exist in the development of a biomarker for use as a diagnostic tool in reflux-associated pulmonary disease. The sensitivity and specificity of salivary pepsin are not sufficient to be considered a highly reliable test in isolation, and collection of samples at only a single time point has been suggested to result in false negatives in patients who are not actively refluxing.9 Therefore, at this time, salivary and BAL pepsin represent one potential tool to aid in the differential diagnosis of chronic pulmonary complaints consistent with laryngeal hyperresponsiveness, and salivary pepsin is particularly appealing, as it does not require specialized equipment or cause patient discomfort. LLMI offers relatively greater sensitivity at the cost of poor specificity, and of course requires the acquisition of BAL specimens, which is prohibitive in many clinical contexts.
PEPSIN AS A MEDIATOR OF INFLAMMATORY DISEASE PROCESSES
In vitro studies have shown that via receptor-mediated endocytosis, nonacid pepsin can enter the epithelium of the hypopharynx and larynx.14, 15 Following endocytosis, receptors and ligands are sorted within weakly acidic late endosomes and the transreticular Golgi (TRG), raising the possibility of pepsin transport via these pathways. Immunoelectron microscopic findings have supported this notion, having identified colocalization of pepsin with the late endosome marker Rab-9 and the TRG marker TRG-46.16 The TRG has a weakly acidic pH of approximately 5, at which pepsin has roughly 40% of its maximal activity17, 18; as such, inactive pepsin might potentially be taken up by laryngeal epithelial cells and be activated within intracellular compartments of low pH, setting the stage for intracellular damage (Figure 5–1). Downstream, exposure of hypopharyngeal cells to pepsin at pH 7 has been shown to induce the expression of several proinflammatory cytokines and receptors, including IL-1α, the neutrophil chemoattractant IL-8, and the eosinophil colony-stimulating factor IL-5.19 Conversely, exposure of laryngeal epithelium to pepsin has been shown to deplete protective proteins such as Sep70 and carbonic anhydrase-III, implying multiple pathways by which pepsin-mediated cell damage might contribute to ongoing inflammation and the endoscopic findings of LPR disease.18 Moreover, the aforementioned proinflammatory cytokine profile, induced in hypoharyngeal tissues independent of acidic refluxate, is similar to that expressed in reflux esophagitis and is known to contribute to ongoing inflammation in the pathophysiology of GERD.19
The above research identifies a novel mechanism by which pepsin might induce cellular injury and inflammation irrespective of the acidity of the extracellular environment, potentially proffering an explanation for the persistence of chronic mucosal inflammation, symptoms, and endoscopic findings in many patients with reflux-attributed laryngeal pathology in spite of therapy with high-dose acid suppression. While pepsin has long been known to play an etiologic role in GERD due to its proteolytic activity in the low-pH environment induced by gastroesophageal reflux episodes, the finding of potentially active intracellular pepsin and induction of a proinflammatory response suggests a role for pepsin in reflux-mediated disease of the airway where pH may be less clinically relevant. The receptor mediated uptake of nonacid pepsin, as can occur following LPR, and any inflammatory or neoplastic changes that may occur as a result,11, 19–21 cannot be prevented by PPIs, which only address acid production in gastric mucosa. As the role of pepsin in LPR-mediated mucosal damage seems to involve its activation within more acidic intracellular compartments or through dysregulation or activation of cell-signaling cascades,16 the amelioration of the acidic environment of gastric refluxate with PPI or histamine (H2 receptor) antagonists may not adequately address pepsin-mediated inflammatory changes.
Figure 5–1. Receptors and their ligands are typically sorted in late endosomes and the TRG. When inactive pepsin is taken up by laryngeal epithelial cells by receptormediated endocytosis, it may be reactivated in these intracellular compartments of lower pH and thereby cause intracellular damage. Alternatively, binding / activation of the cell surface receptor may induce a cellsignaling event that negatively affects cell function.
Although PPIs remain the mainstay for treatment of GERD, there is poor evidence for their efficacy in the treatment of airway reflux-mediated disease, including LPR.22 It is widely believed that the upper airway is more sensitive to reflux than the esophagus, and therefore higher dose PPIs are necessary for the control of LPR-related symptoms.23–25 At this time, placebo-controlled studies by and large have not shown a significant therapeutic benefit to PPIs used in LPR.26–31 Although some studies have noted evidence of symptomatic improvement with PPI therapy,32, 33 upon review of these two studies, it has been argued that the affected patients only had significant improvement of gastroesophageal reflux symptoms rather than improvement of upper airway symptoms.30 Arguments can be made that these studies were done prior to the era of HEMII-pH testing, and that the diagnosis of nonacid reflux was incomplete. In light of the poor data for the efficacy of acid suppression in treatment of extraesophageal reflux, the American Gastroenterological Association has specifically recommended against the empiric use of PPIs for suspected LPR unless there are concomitant symptoms of GERD.34 Likely as a result of the paucity of alternative effective therapies, however, PPIs continue to be used for LPR,35 and indeed the American Academy of Otolaryngology—Head and Neck Surgery has recommended empiric use of high-dose PPI therapy for suspected LPR, with laparoscopic fundoplication proposed as an alternative to medical management.24 A recent survey from the American Bronchoesophagological Association reported that twice-daily PPIs remain a popular first-line therapy for LPR.36
Laparoscopic fundoplication and magnetic ring procedures are well-established, reliable options for the surgical management of GERD. In contrast to the predictable improvement seen in the treatment of GERD, research on the efficacy of antireflux surgery in the treatment of LPR is mixed, with various studies showing resolution of chronic cough ranging from 63% up to 85% of patients.37–39 Hypotheses for this variance range from differences in surgical technique to differences in patient selection criteria. In particular, it has been observed that patients with more severe stereotypical GERD symptoms are more likely to benefit from antireflux surgery,37, 40 and in particular patients with preoperative heartburn and pH <4 for more than 12% of a 24-hour period have been found to have a 90% probability of symptomatic improvement.40
The role of nonacid reflux in LPR is additionally supported by findings using MII-pH monitoring, which represents a major advance in the diagnosis of extraesophageal reflux disease due to its ability to detect nonacidic reflux events. MII-pH has demonstrated a strong association of LPR symptoms,41 which, combined with the previously described research revealing poor evidence for the efficacy of acid suppression in LPR and the substantial improvement of symptomatology following surgical intervention such as fundoplication that prevents refluxate from reaching the laryngopharynx,12, 40, 42, 43 suggests the nonacid components of refluxate must play a major role in the pathophysiology of laryngopharyngeal reflux with chronic cough. In the absence of clear consensus or guidelines regarding the appropriateness of antireflux surgery, preoperative evaluation of esophageal reflux events and motility with MII-pH and high-resolution esophageal manometry has been suggested as a useful tool in predicting postoperative symptom improvement.12, 43 These smaller studies refute the claim that GERD alone predicts response to Nissen by utilizing the newest HEMII-pH technology to select candidates for ant-reflux surgery.12, 43
Prior in vitro studies have also suggested that bile can cause laryngeal inflammation irrespective of pH, but others have countered that no evidence exists for this as a mechanism of injury to the human larynx.44 While in vitro exposure to bile acids causes “blebbing” of cell membranes,45 to our knowledge there are no reports of this histologic finding in laryngeal mucosa of patients with LPR. Notably, previous studies have used high concentrations of bile salts and acids, ranging from 5 to 50 mM. This overlaps the physiologic concentration of bile salts in the human duodenum, which ranges from 10 to 22 mM.46 In contrast, the physiological bile acid/salt content in gastric refluxate reaching the laryngopharynx is expected to be in the micromolar range,47 which to our knowledge has not been known to cause damage to laryngeal mucosa. Further, unconjugated bile acids, which cause damage at higher pH such as that of the laryngopharynx, are rarely found as components of gastric refluxate.11, 47 However, there is no strong evidence for the concentration of each bile acid within the laryngopharynx, and the role of bile acids in the pathophysiology of LPR thus remains a dilemma. As such, the previous prevailing understanding of LPR and airway reflux as acid-mediated disease have shifted to a model in which disease is largely mediated by nonacid components of refluxate in which traditional medical management with acid suppression is not sufficient.48
As discussed above, PPIs continue to be commonly used in clinical practice for the treatment of airway reflux disease including LPR in spite of poor evidence for their efficacy,26–31 with approximately $26 billion spent yearly for this indication.49 In light of the inefficacy of PPI therapy for LPR and its associated costs and potential risks, there is substantial interest in an alternative modality for the treatment of LPR.48, 50, 51
Pepsin represents an exciting potential novel target for future therapies, particularly for patients who experience symptoms refractory to PPI therapy in light of its role in nonacid LPR.16, 48 Prior literature has identified two mechanisms by which pepsin might be targeted in therapy of LPR: by irreversible inactivation and via receptor antagonism.16, 48 While the first of these would prevent pepsin’s reactivation within the acidic environment of intracellular compartments (Figure 5–2A), the latter would prevent its endocytosis and perhaps the previously described induction of expression of a proinflammatory cytokine profile (Figure 5–2B). Although Pepstatin A is a potent inhibitor of pepsin activity and is currently commercially available, its poor pharmacokinetics and water-soluble characteristics make it a relatively poor candidate for the purpose of treating LPR. As such, novel agents targeting pepsin are currently in development and represent an exciting potential avenue for the treatment of reflux-mediated disease including LPR and chronic cough.
THINKING OUTSIDE OF THE BOX
Although the above literature identifies a clear role for pepsin in the pathophysiology of LPR and chronic cough, studies investigating the utility of salivary pepsin as a marker of extraesophageal reflux found mixed results. Although this may suggest pepsin is not a useful biomarker for airway disease, it is important to remember that salivary levels of pepsin may only be transiently elevated as a result of the transient nature of reflux and intermittent influence of swallowing, both of which will vary with food intake. To that end, optimization of the timing and method of acquisition of salivary samples to yield the greatest balance of sensitivity and specificity is essential if pepsin is to be used clinically as a marker of disease. Although salivary and sputum samples are typically collected at least 1 hour after meals to avoid detection of postprandial reflux events, there have not been any large-scale studies to determine the optimal timing or method of sample collection, and hence the optimal timing for collection of samples is still unclear. Larger studies with more frequent collection of samples may allow determination of the ideal time for sample collection, which might in turn facilitate the use of salivary pepsin as a more sensitive and specific tool for diagnosing airway disease. Furthermore, salivary pepsin may help to identify a specific subset of chronic cough patients whose reflux and aspiration contributes to their disease, and thus a subset of patients who might benefit from lifestyle modifications for reflux and surgery.
While salivary pepsin does represent evidence of reflux events, and tracheal or BAL pepsin is clear evidence of aspiration, measurement of pepsin in secretions may not adequately reflect its uptake via receptor-mediated endocytosis in laryngeal and lung epithelial cells. As such, intracellular pepsin and tissue-bound pepsin may be more indicative of pepsin-mediated disease than salivary or tracheal/BAL pepsin. Other factors, such as compromise of mucosal defense mechanisms of the tissues in question, may predispose to increased endocytosis of pepsin, setting the stage for its intracellular activation and downstream development of an inflammatory response. Although the invasiveness of sample collection would likely preclude the use of pepsin within tissues as a diagnostic marker, the precise role of pepsin in LPR and chronic cough may be elucidated by future studies examining both factors, which might contribute to increased receptor-mediated endocytosis of pepsin and the relationship between pepsin in salivary secretions and pepsin in laryngeal tissues.