Chronic Cough Due to Gastroesophageal Reflux Disease...

DOI 10.1378/chest.129.1_suppl.80S 2006;129;80S-94SChest

Richard S. Irwin Clinical Practice GuidelinesReflux Disease : ACCP Evidence-Based Chronic Cough Due to Gastroesophageal

http://chestjournal.chestpubs.org/content/129/1_suppl/80S.full.html

services can be found online on the World Wide Web at: The online version of this article, along with updated information and

ISSN:0012-3692)http://chestjournal.chestpubs.org/site/misc/reprints.xhtml(

written permission of the copyright holder.this article or PDF may be reproduced or distributed without the priorDundee Road, Northbrook, IL 60062. All rights reserved. No part of Copyright2006by the American College of Chest Physicians, 3300Physicians. It has been published monthly since 1935.

is the official journal of the American College of ChestChest

© 2006 American College of Chest Physicians by guest on July 18, 2011chestjournal.chestpubs.orgDownloaded from

http://chestjournal.chestpubs.org/content/129/1_suppl/80S.full.html

http://chestjournal.chestpubs.org/site/misc/reprints.xhtml

http://chestjournal.chestpubs.org/

Chronic Cough Due toGastroesophageal Reflux DiseaseACCP Evidence-Based Clinical Practice Guidelines

Richard S. Irwin, MD, FCCP

Objectives: To critically review and summarize the literature on cough and gastroesophagealreflux disease (GERD), and to make evidence-based recommendations regarding the diagnosisand treatment of chronic cough due to GERD.Design/methodology: Ovid MEDLINE literature review (through March 2004) for all studiespublished in the English language and selected articles published in other languages such asFrench since 1963 using the medical subject heading terms “cough,” “gastroesophageal reflux,”and “gastroesophageal reflux disease.”Results: GERD, singly or in combination with other conditions, is one of the most common causesof chronic cough. In patients with normal chest radiographic findings, GERD most likely causescough by stimulation of an esophageal-bronchial reflex. When GERD causes cough, there may beno GI symptoms up to 75% of the time. While 24-h esophageal pH monitoring is the most sensitiveand specific test in linking GERD and cough in a cause-effect relationship, it has its limitations.In addition, there is no general agreement on how to best interpret the test, and it cannot detectnon-acid reflux events. Therefore, when patients fit the clinical profile that has a high likelihoodof predicting that GERD is the cause of cough, antireflux medical therapy should be empiricallyinstituted. While some patients improve with minimal medical therapy, others require moreintensive regimens. When empiric treatment fails, it cannot be assumed that GERD has beenruled out as a cause of chronic cough. Rather, an objective investigation for GERD is thenrecommended because the empiric therapy may not have been intensive enough or medicaltherapy may have failed. Surgery may be efficacious when intensive medical therapy has failed inselected patients who have undergone an extensive objective GERD evaluation.Conclusions: Accurately diagnosing and successfully treating chronic cough due to GERD can bea major challenge. (CHEST 2006; 129:80S–94S)

Key words: acid reflux disease; clinical profile of cough due to gastroesophageal reflux disease; cough due togastroesophageal reflux disease; esophageal-bronchial cough reflex; gastroesophageal reflux; gastroesophageal refluxdisease; macroaspiration; microaspiration; nonacid reflux disease; silent reflux disease

Abbreviations: GER � gastroesophageal reflux; GERD � gastroesophageal reflux disease; PPI � proton pump inhib-itor; BAL � bronchoalveolar lavage

T he origin of the terms gastroesophageal reflux(GER) and gastroesophageal reflux disease

(GERD) is Latin.1 Reflux literally means backflow,coming from the two Latin words (re meaning“back,” plus fluere meaning “to flow”). It follows that

GER is a term that refers to the backflow of stomachcontents into the esophagus. Because GER episodescan occur up to 50 times a day, usually during mealsand the postprandial state in healthy individuals, andnot produce any symptoms, GER events can bephysiologic in nature.2 On the other hand, GERD isa term that refers to a disease that is caused by GERmanifested by either symptoms and/or tissue dam-age.

GERD is a term that encompasses a spectrum ofcommon clinical manifestations from localized

Reproduction of this article is prohibited without written permissionfrom the American College of Chest Physicians (www.chestjournal.org/misc/reprints.shtml).Correspondence to: Richard S. Irwin, MD, FCCP, University ofMassachusetts Medical School, Room S6-842, 55 Lake Ave North,Worcester, MA; e-mail address: [email protected]

80S Diagnosis and Management of Cough: ACCP Guidelines



esophageal manifestations with typical and promi-nent GI symptoms, such as heartburn and regurgi-tation, to extraesophageal problems, such as cough orhoarseness, without any GI complaints. For example,in a population-based study in Olmsted County, MN,Locke et al,3 using a validated GER questionnaire,identified a prevalence rate of 59% for occasionalepisodes of GER-related symptoms such as heart-burn, regurgitation, and dysphagia; 20% of the re-spondents reported weekly symptoms. In a prospec-tive before-and-after intervention trial, Irwin et al4found that when GERD was the cause of chroniccough, it could be silent from a GI standpoint up to75% of the time. In a study in which the symptomsand findings of GERD were specifically sought andprospectively compared in otolaryngology and gas-troenterology patients, Ossakow et al5 found thathoarseness was present in 100% of the otolaryngol-ogy patients and 0% of the gastroenterology patients,while heartburn was present in 89% of the gastroen-terology patients and in only 6% of the otolaryngol-ogy patients. When otolaryngologists diagnose laryn-geal or pharyngeal disease due to the backflow ofstomach contents into the laryngopharynx, they referto it as laryngopharyngeal reflux disease.1

It is now clear that the spectrum of clinicalmanifestations of GERD as evaluated by differentspecialists can vary.1,4,6 It is also becoming increas-ingly more likely that the management of a variety ofextraesophageal manifestations of GERD cannotonly differ in clinical and diagnostic ways from theclassic GERD seen by gastroenterologists, but alsoperhaps in pathogenetic and therapeutic aspects aswell. Physicians who manage patients with coughshould be aware of these differences. While there ismuch that has been learned about a variety ofesophageal and extraesophageal manifestations ofGERD, this section will focus solely on chroniccough. Literature on screening for and managingcomplications of esophageal GERD, such as Barrettepithelium and esophageal cancer, can be foundelsewhere.7,8 This section reflects an Ovid MED-LINE literature review (through March 2004) for allstudies published in the English language and se-lected articles published in other languages such asFrench since 1963 using the medical subject headingterms “cough,” “gastroesophageal reflux,” and “gas-troesophageal reflux disease.”

Prevalence of GERD as a Cause of Cough

When the diagnosis of chronic cough due toGERD is based on a favorable response (eg, elimi-nation of or improvement in cough) to specificGERD treatment, prospective before-and-after in-

tervention studies9–14 have revealed that GERD,singly or in combination with other conditions, is oneof the most common causes of chronic cough inadults in the world. The prevalence in these studiesranged from 5 to 41%. While it is not clear whataccounts for the variation in prevalence, multiplefactors likely contribute to it, such as differences inthe populations referred and studied. For example,in three prospective studies published by Irwin etal9,15,16 over a period of 17 years, the prevalence andrelative commonness of GERD as the cause ofchronic cough increased from a low of 10% andbeing the fourth most common cause in 1981,15 to21% and the third most common cause in 1990,9 to36% and the second most common cause in 1998.16

Pathophysiology

Knowledge of what is known about where and howGERD can cause cough provides a framework forunderstanding what has been learned about diagno-sis and treatment.17 Although GERD can stimulatethe afferent limb of the cough reflex by irritating theupper respiratory tract without aspiration (eg, thelarynx) and by irritating the lower respiratory tract bymicroaspiration or macroaspiration, there is evi-dence from a randomized and controlled study byIng and colleagues18 that strongly suggests thatGERD also can cause chronic cough by stimulatingan esophageal-bronchial cough reflex. By this neuralreflex mechanism, refluxate into the distal esophagusalone is thought to be a sufficient stimulus to causecough.19

When GERD causes cough by irritating the lar-ynx, laryngoscopy has the potential to demonstratesigns consistent with “reflux laryngitis” (eg, posteriorlaryngitis with red arytenoids and piled-up interary-tenoid mucosa).20 Bronchoscopy and chest-imagingstudies have the potential to detect abnormalitiesconsistent with aspiration. Bronchoscopy may revealairway signs consistent with aspiration (eg, subglotticstenosis,2 hemorrhagic tracheobronchitis, and ery-thema of subsegmental bronchi);21 chest-imagingstudies may demonstrate a wide spectrum of paren-chymal abnormalities.22 While laryngoscopic andbronchoscopic signs are consistent with reflux-in-duced injury, one must be cautious and not assumethat merely observing changes consistent with in-flammation and edema of the larynx and lowerairways is specific for GERD because they may alsopotentially be due to the act of coughing itself that isprovoked by other diseases.23 When laryngoscopy,bronchoscopy, and chest-imaging studies are normal,it is assumed that GERD has caused cough bystimulating the esophageal-bronchial reflex. Multiple

www.chestjournal.org CHEST / 129 / 1 / JANUARY, 2006 SUPPLEMENT 81S



prospective clinical studies17 have suggested thislatter pathophysiologic scenario to be the most com-mon.

Evidence is mounting (eg, prospective before-and-after, successful antireflux surgery intervention trialin patients who had previously failed to improvedespite total/near-total elimination of esophagealacid24; prospective, randomized, controlled pharma-cologic studies18; and randomized, controlled intra-esophageal acid/saline solution challenge studies4,18)that there are likely to be multiple potential triggersof cough including, but not limited to, acid in thegastric refluxate, and the triggers of cough may differamong patients. While it is not known what thenonacid factors are, they may include4,24 alkaline pH,pancreatic enzymes, bile, and esophageal dysmotil-ity.25,26 Because esophageal dysmotility can be com-monly demonstrated in patients with GERD, it is notclear whether or when esophageal dysmotility is anadverse occurrence due to GERD or contributing toGERD. Because the term acid reflux disease whenapplied to chronic cough due to GERD can be amisnomer and may mislead clinicians into thinkingthat all patients with cough due to GERD shouldimprove with acid-suppression therapy, it is recom-mended that the term acid reflux disease no longerbe used in the context of cough unless it can bedefinitively shown to apply. Unless acid reflux dis-ease can be definitively proven, the more generalterm reflux disease should be used.

Because coughing can induce GER episodes,27 acough-GER self-perpetuating cycle may be involvedin the pathophysiology of a patient’s chroniccough.18,28 While cough-induced GER episodes areoften observed in 24-h esophageal pH-monitoringtracings, the mechanisms by which these GER epi-sodes are triggered by coughing is not known.

Based on inhaled tussigenic cough challenges,there is little evidence to support the theory that anincreased sensitivity of the cough reflex is the soleexplanation why some patients with GERD complainonly or predominantly of cough, while others primar-ily have GI complaints. An increased sensitivity hasbeen observed in patients with GERD who do notcough29 as well as in those who do cough.30

Recommendation

1. In patients with chronic cough due toGERD, the term acid reflux disease, unless itcan be definitively shown to apply, should bereplaced by the more general term reflux dis-ease so as not to mislead the clinicians intothinking that all patients with cough due to

GERD should improve with acid-suppressiontherapy. Level of evidence, expert opinion; benefit,substantial; grade of recommendation, E/A

Diagnosis

Clinical Presentation

There is nothing about the character and timing ofthe cough due to GERD that distinguishes it fromother causes of cough.31 It can present as a cough-phlegm (ie, productive cough) syndrome, just likechronic bronchitis from cigarette smoking, as well asa dry cough. It occurs nocturnally in only a minorityof patients,31 and it can be “silent” from a GIstandpoint up to 75% of the time.4

On the other hand, GERD should always beconsidered as a possible cause of chronic coughwhen patients also complain of typical and frequentGI symptoms such as daily heartburn and regurgita-tion, especially when the chest-imaging studiesand/or clinical syndrome are consistent with anaspiration syndrome.22 Aspiration syndromes associ-ated with cough due to GERD include Mendelsonsyndrome, bacterial pneumonia and lung abscess,chemical pneumonitis, exogenous lipoid pneumonia,recurrent bacterial pneumonias, chronic interstitialfibrosis, bronchiectasis, Mycobacterium fortuitum orMycobacterium chelonei pneumonia, diffuse aspira-tion bronchiolitis, and tracheobronchitis.22

While the character and timing of cough are notpredictive of when GERD is likely to be the cause ofchronic cough, there is a clinical profile that has beenhighly predictive (approximately 91%) that the pa-tient’s cough will respond to antireflux treatment,even when there are no GI complaints. This clinicalprofile, which appears in Table 1, was initially sug-

Table 1—Clinical Profile That Predicts That ChronicCough Is Likely Due to GERD

Chronic coughNot exposed to environmental irritants nor a present smokerNot taking an angiotensin-converting enzyme inhibitorChest radiograph is normal or shows nothing more than stable,

inconsequential scarringSymptomatic asthma has been ruled out:

Cough has not improved with asthma therapy, orMethacholine inhalation challenge is negative

Upper airway cough syndrome due to rhinosinus diseases has beenruled out:

First-generation H1-antagonist has been used and cough failedto improve, and

“Silent” sinusitis has been ruled outNonasthmatic eosinophilic bronchitis has been ruled out:

Properly performed sputum studies are negative, orCough has not improved with inhaled/systemic corticosteroids




gested by post hoc analyses of four prospective,before-and-after intervention studies9,15,31,32 of pa-tients with chronic cough due to a variety of diseases,and was prospectively assessed in two small, before-and-after intervention studies4,33 of patients withchronic cough due to GERD. In one of theseprospective studies,4 cough disappeared in 12 of 12patients receiving intensive medical therapy; in theother study,33 cough disappeared or significantlyimproved in 18 of 21 patients who had undergoneantireflux surgery after they had not responded tointensive medical therapy.

The results of a prospective, before-and-after in-tervention trial24 have suggested that GERD cannotbe excluded on clinical grounds as the potentialcause of cough. This study suggested that there canbe a dissociation in improvement in cough and GIcomplaints. Chronic cough due to GERD in somepatients remained unchanged, while GI symptomsimproved or disappeared with medical antirefluxtherapy. Moreover, because prior antireflux surgeryfailed to control GERD in some of the patients,perhaps for technical reasons, this study also sug-gested that it cannot be assumed that GERD hasbeen definitively ruled out as a cause of coughbecause there is a history of prior antireflux surgery.

A number of potential risk factors for GERDhave been identified. These include the following:drugs such as alendronate,34 oral corticosteroids,35

bronchodilators such as inhaled �2-adrenergicagonists36 and theophylline in some studies,37

progesterone,37 calcium channel-blockingagents,37anticholinergic agents,37 morphine,37 andmeperidine37; obesity38; smoking37; vigorous exer-cise39,40; alcohol41; caffeine in some studies37; fattyfoods, chocolate, carminatives, and irritating foodssuch as citrus juices and tomato products37; pro-longed gastric intubation42,43; lung transplanta-tion44; following pneumonectomy45 and peritonealdialysis46,47; and a variety of respiratory diseasessuch as asthma48 and obstructive sleep apnea.49 –51

While it is not known what their presence meansfrom a diagnostic standpoint in the evaluation ofcough, the failure to correct or eliminate riskfactors may assume importance during treatment.This will be discussed in the “Treatment” section.Based on a small genome-wide scan52 of familiesaffected by severe pediatric GERD, followed byhaplotying and by pairwise and multipoint linkageanalyses based on epidemiologic studies, it ispossible that GERD, in the future, may one day befound to be a genetically determined disease, andthat genetic testing may be able to predict who isat risk for GERD and its varied presentations.52

Laboratory Testing

Depending on the information that is sought,there are a variety of laboratory tests from which tochoose. For example, the results of esophageal ma-nometry to assess for a hypotensive lower esophagealsphincter and barium esophagography to assess forthe presence of a hiatus hernia may suggest thatthere is the potential for GERD. While bariumesophagography results may suggest that there ismucosal injury, esophagoscopy with mucosal biopsycan confirm it. Intraesophageal hydrochloric acidinfusion while monitoring symptoms (ie, the Bern-stein test) may reveal esophageal mucosal acid sen-sitivity. Barium esophagography, radionuclideesophageal scintigraphy, and esophageal pH moni-toring can assess for GER events. Dual esophagealand pharyngeal simultaneous pH monitoring canevaluate for gastropharyngeal GER events and theirpotential to cause laryngopharyngeal injury, and therisk of aspiration. A radionuclide gastric-emptyingstudy with solids has the potential to assess forGERD that may be contributed to by delayed gastricemptying. Last, 24-h esophageal pH monitoring,lung scintiscanning after intragastric 99mTc-sulfurcolloid infusion, investigating for lipid-laden alveolarmacrophages in induced sputum or BAL fluid andobserving the gross appearance of airway mucosalsurfaces during laryngoscopy and bronchoscopy, andlung biopsy have been utilized to link GER with avariety of pulmonary diseases as well as cough.

While tests that link GER with cough suggest apotential cause-effect relationship, a definitive diag-nosis of cough due to GERD requires that coughnearly or completely disappear with antireflux treat-ment.28 Utilizing a favorable response to specificantitussive therapy as the “gold standard” test forcomparison, and following previously published evi-dence-based guidelines for interpreting 24-h esoph-ageal pH monitoring and treating GERD in thecontext of chronic cough,28 three prospective before-and-after intervention trials4,9,32 have revealed thefollowing.17 Although the standard catheter-based24-h esophageal pH monitoring has been shown tobe the single best test to link GERD and cough in acause-effect relationship, because it is the mostsensitive and specific test, it is important to be awarethat conventionally utilized diagnostic indexes ofGERD (eg, percentage of time that pH is � 4) canbe misleadingly normal and that observing GER-induced coughs can more often be helpful. Conse-quently, the sensitivity of the test can depend on howit is interpreted. Until future studies provide betterguidelines, it is recommended that the test be inter-preted as normal only when conventional indexes foracid reflux are within the normal range and no




reflux-induced coughs appear during the monitoringstudy. For example, in three prospective stud-ies,4,33,53 a cough was considered to be induced byacid reflux only when cough occurred simultaneouslywith or within 3 min following a drop in pH to � 4.These criteria were empirically established and willneed to be validated by future prospective studies.When cough is due to GERD, it should not beexpected that there will necessarily be a high per-centage of coughs that appear to be induced by GERevents in the 24-h esophageal pH tracing. In twoprospective, before-and-after intervention trials inwhich it was determined that chronic cough was dueto GERD, GER events in the distal esophagusappeared to induce an average of 29% of the pa-tients’ coughs (range, 0 to 100 coughs)4 and 35% ofthe patients’ coughs (range, 0 to 89 coughs).53 It isalso important to stress that the degree of abnormal-ity noted in the esophageal pH-monitoring variablesdoes not directly correlate with the severity of thepatients’ cough.54

The standard, catheter-based 24-h esophageal pH-monitoring study has its own inherent limitations.17

For example, because it cannot detect reflux eventswith a pH similar to that of the normal esophagus,the monitoring study can be entirely normal at a timewhen nonacid GERD is the cause of cough. In thissituation, barium esophagography may be the onlyavailable test to reveal GER of potential pathologicsignificance (see the discussion below regardingesophageal impedance monitoring). Barium esopha-gography has revealed reflux to the mid-esophagusor higher when refluxate from the stomach had a pHvalue similar to that of the normal esophagus.4Because the majority of patients with cough due toGERD do not have esophagitis or Barrett epitheli-um,4 a normal esophagoscopy finding does not ruleout GERD as the cause of cough. While pilotstudies55 have shown that ambulatory esophageal pHmonitoring using a wireless recording system is aviable option for patients who are unwilling orunable to undergo conventional ambulatory pH-monitoring studies using a transnasally positionedpH catheter, it is likely that future studies will showthat the wireless system will have the same limita-tions as the standard catheter-based systems, whichwere summarized above.

Because 24-h esophageal pH monitoring in thecontext of prospectively evaluating chronic cough hasa sensitivity of � 100% (approximately 90%) and itsspecificity has ranged between 66% and 100%,4,9,32

even when interpreted according to the guidelinesdescribed above, and because these ranges havebeen confirmed in a retrospective study,56 it isrecommended that treatment be initially started inlieu of testing for patients who fit the clinical profile

for cough due to GERD (Table 1). This recommen-dation is supported by a retrospective, before-and-after intervention study57 in which GERD was diag-nosed as the cause of chronic cough in 79% ofpatients (44 of 56 patients) by the empiric trial oftherapy, thus obviating the need to perform 24-hesophageal pH monitoring. The performance of 24-hesophageal pH monitoring is recommended duringtherapy when cough does not improve or resolve toassist in determining whether the therapy needs tobe intensified or medical therapy has failed. Theusefulness of this recommendation was shown in asmall prospective before-and-after intervention tri-al.24 Hopefully, in the near future, it will be deter-mined whether simultaneous 24-h monitoring ofintraesophageal impedance and pH will allow us tomore easily determine when acid reflux diseaseand/or nonacid reflux disease are the causes ofchronic cough.58,59

While impedance monitoring allows for the detec-tion of air and liquid reflux without requiring a pHshift to detect the liquid reflux, the simultaneousmonitoring of pH helps to determine whether theliquid reflux observed by impedance was acidic innature. Dual-impedance pH monitoring may alsohelp to determine the roles of esophageal manome-try in diagnosing cough and esophageal dysmotilityin causing cough due to GERD.25,26,60 While a highprevalence of abnormal esophageal manometricmeasurements have been reported in a group ofpatients with chronic cough thought to be due toGERD, and manometry was the only abnormal testresult in 32% of patients,60 a randomized, prospec-tive study utilizing a control group of patients withcough due to conditions other than GERD will berequired to determine whether manometry is auseful diagnostic test for distinguishing cough due toGERD from cough due to other diseases. Moreover,a randomized, prospective, double-blind, placebo-controlled trial will be required to determinewhether esophageal dysmotility is pathogeneticallylinked to cough due to GERD.

As stated above, the findings of chest-imagingstudies, laryngoscopy, and bronchoscopy can be en-tirely normal when GERD is the cause of cough. Inthis context, GERD most likely has caused cough viastimulation of an esophageal-bronchial reflex path-way.28 While laryngoscopic findings61 and broncho-scopic findings,2,21 putatively due to GERD, havebeen published, it is not known how specific any ofthe findings are for injury due to GERD becausethey may also be potentially caused by the trauma ofcoughing itself23 (also see the “Complications” sec-tion elsewhere in this Supplement).

Lipid-laden alveolar macrophages have been pro-spectively evaluated as a marker of aspiration by




BAL in patients with pulmonary parenchymal lungdiseases62 and by induced sputum in patients witheither GI symptoms without respiratory complaintsor patients with mild asthma in the context of normalchest radiographs.63 In the first study,62 the merepresence of lipid-laden alveolar macrophages inlower respiratory tract secretions was found to be anonspecific marker of parenchymal lung diseasebecause the presence and/or appearance of intracel-lular lipid in these cells was not helpful in differen-tiating aspiration from nonaspiration as the cause ofthe lung disease. A post hoc analysis of the datarevealed that observing few or no lipid-laden alveolarmacrophages might be helpful in ruling out aspira-tion as the cause of the parenchymal lung disease.This conclusion was reached on the basis of thefollowing information: a lipid cellular index of � 100,which was computed by multiplying the observedamount of intracellular lipid in 100 cells, gradedfrom 0 to 4, had a specificity of 57%, and sensitivityand negative predictive values of 100%. In thesecond study,63 while it was found that the presenceof an increased number of lipid-laden macrophagesin induced sputum predicted the presence of oro-pharyngeal reflux by dual-channel 24-h esophagealand pharyngeal pH monitoring, it was not deter-mined that these same patients actually had aspi-rated. Most relevant to the discussion in this section,the diagnostic usefulness of the presence of lipid-laden macrophages in induced sputum was notassessed in patients with chronic cough. On the basisof these studies, assessing for lipid-laden macro-phages in BAL fluid and induced sputum does notappear, at this time, to have a role to play indiagnosing cough due to GERD.

Because inhaled tussigenic challenges with capsa-icin have revealed the increased sensitivity of thecough reflex in a variety of diseases (eg, asthma, viralupper respiratory tract infections, and angiotensin-converting enzyme inhibitors) in addition toGERD30 and even in patients with GERD withoutcough,29 inhaled tussigenic challenges with capsaicinat this time do not appear to have a role to play indiagnosing cough due to GERD. Because exhalednitric oxide can be a marker of airway inflammationin patients with asthma, its measurement may be ofpotential benefit in diagnosing the cause of chroniccough. However, preliminary studies64,65 to datehave not shown exhaled nitric oxide measurementsto be particularly helpful in diagnosing cough due toGERD. In a cross-sectional prospective study65 uti-lizing exhaled nitric oxide as well as induced sputum,it was determined that GERD, when associated withcough or mildly symptomatic asthma, did not causeor aggravate existing airway inflammation as mea-

sured by exhaled nitric oxide or other indexes ofairway inflammation (eg, cell counts and fibrinogenlevel).

Because two prospective studies4,18 in subjectswith chronic cough due to GERD have produceddivergent results during intraesophageal infusions ofhydrochloric acid to provoke cough, a negative Bern-stein test cannot be used to exclude the diagnosis ofcough due to GERD. While a radioisotope scintiscanthat documents pulmonary aspiration of gastric con-tents is definitive, it is not routinely performedbecause it may be positive in not � 50% of subjectswith suspected aspiration, and it has a much lowersensitivity.

Empiric Trial of Therapy

The first consensus panel report28 in 1998 recom-mended that, when 24-h esophageal pH monitoringcannot be performed or is not available, a diagnostic,empiric trial of medical antireflux therapy be per-formed in patients who meet the clinical profile(Table 1) predicting that silent GERD is the likelycause of chronic cough or in patients with chroniccough who also have prominent upper GI symptomsthat are consistent with GERD. The first consensuspanel report28 also recommended that if empirictreatment fails, it cannot be assumed that GERD hasbeen ruled out as a cause of chronic cough; rather,objective investigation for GERD is then recom-mended because the empiric therapy may not havebeen intensive enough or medical therapy may havefailed.

Since 1998, a number of articles have been pub-lished that support the above two recommendations.Poe and Kallay,57 in a retrospective before-and-afterempirical medical intervention trial, provided datasuggesting that an empiric trial of therapy can besuccessful; they were able to avoid the performanceof 24-h esophageal pH monitoring in 44 of 56patients in whom cough due to GERD had beendetermined. Others have provided data that supportthe contention that a negative finding in an empirictrial of therapy does not rule out GERD as a cause ofcough. Peghini et al,66 in a controlled investigationthat assessed for nocturnal acid breakthrough inpatients with GI complaints who were receivingomeprazole (20 mg bid) or lansoprazole (30 mg bid)and in healthy volunteers, showed that nocturnalacid breakthrough occurred in the majority of allsubjects. This study underscores the importance ofbeing aware that a fixed dose of medication may notbe adequate in all patients. Allen and Anvari,67 in aprospective before-and-after antireflux surgery inter-vention trial in patients who had failed to improvewith intensive medical therapy, showed that laparo-




scopic reflux surgery had improved or cured coughin 85% of chronic coughers 6 months after surgery.Novitsky et al,33 in a prospective before-and-afterantireflux surgery intervention trial in patients withchronic cough who had failed to respond to intensivemedical therapy, reported an 86% improvement ofchronic cough 12 months following surgery. Irwin etal,24 in a small, prospective, before-and-after inter-vention trial in patients with chronic cough who hadfailed to respond to very intensive medical therapy,reported the improvement or elimination of cough inall subjects 12 months following surgery. In thisstudy, cough persisted prior to surgery even thoughthe antireflux medical regimen was serially intensi-fied until there was total or near-total elimination ofesophageal acid on repeated 24-h esophageal pHmonitoring. This finding supported the diagnosis ofnonacid reflux disease.

The results of a metaanalysis68 of 15 trials (open-label, 2 trials; randomized, controlled, 10 trials;randomized, controlled, crossover, 3 trials) that didnot involve extraesophageal symptoms such as coughshowed that the short-term treatment (ie, 1 to 4weeks) of GERD with normal-dose or high-doseproton pump inhibition (ie, the so-called PPI test) inpatients who were suspected of having GERD didnot confidently establish or exclude the diagnosiswhen GERD was defined by objective referencestandards. In patients with GI symptoms suggestiveof GERD in which the clinical response to a shortcourse of proton pump inhibition was compared with24-h esophageal pH monitoring, the positive likeli-hood of a symptomatic response detecting GERDranged from 1.63 to 1.87. The sensitivity was 0.78(95% confidence interval, 0.66 to 0.86) and specific-ity 0.54 (95% confidence interval, 0.44 to 0.65).

While Ours et al69 concluded from their prospec-tive, double-blind, randomized, placebo-controlledstudy that the best diagnostic and therapeutic ap-proach to chronic cough is a fixed dose of omepra-zole, 40 mg bid for 2 weeks, there are many meth-odological concerns about this study that make theresults uninterpretable.70 With respect to concernsabout GERD, there were 17 patients with chroniccough who had abnormal findings of 24-h esophagealpH-monitoring studies, and only 6 patients improvedwith omeprazole therapy for a response rate of 35%.Because the authors assumed that the dose andduration of omeprazole therapy that they used wouldcure all patients with cough due to GERD, theyreasoned that the other patients with abnormalfindings of 24-h esophageal pH-monitoring studiescould not have GERD. In the patients who did notrespond, the authors did not (1) determine whethertheir medical therapy was maximal, (2) consider thattheir subjects might have had nonacid reflux disease

and therefore required more than acid suppressiontherapy, and (3) did not find another cause for thechronic coughs of the patients.

On the basis of the above considerations, the panelrecommends that an empiric trial of medical antire-flux therapy be performed in patients who meet theclinical profile (Table 1) predicting that silent GERDis the likely cause of chronic cough or in patientswith chronic cough who also have prominent upperGI symptoms (eg, heartburn, sour taste, and regur-gitation) that are consistent with GERD. The paneldoes not think that it is necessary to order tests toassess for the potential of GERD before observingthe response to empiric therapy. While it is notknown what constitutes the minimum effective med-ical therapy for treating the majority of patients withchronic cough due to GERD, the panel, by expertopinion, recommends (1) dietary and lifestyle mod-ifications, (2) acid suppression therapy, (3) the addi-tion of prokinetic therapy (ie, to enhance gut motil-ity) either initially or if there is no response to thetherapy stated in steps 1 and 257 (see section onTreatment), and (4) that the response to therapy beassessed within 1 to 3 months. In some patients,there can be a delay of 2 to 3 months in improvementwith therapy that will eventually eliminate thecough.53 If empiric treatment fails, it cannot beassumed that GERD has been ruled out as a cause ofchronic cough; rather, the objective investigation forGERD is then recommended because the empirictherapy may not have been intensive enough ormedical therapy may have failed.

Recommendations

2. In patients with chronic cough who alsocomplain of typical and frequent GI complaintssuch as daily heartburn and regurgitation, es-pecially when the findings of chest-imagingstudies and/or clinical syndrome are consistentwith an aspiration syndrome, the diagnosticevaluation should always include GERD as apossible cause. Level of evidence, low; benefit,substantial; grade of recommendation, B

3. Patients with chronic cough who have GIsymptoms that are consistent with GERD orwho fit the clinical profile described in Table 1,should be considered to have a high likelihoodof having GERD and should be prescribedantireflux treatment even when they have no GIsymptoms. Level of evidence, low; benefit, substan-tial; grade of recommendation, B

4. In patients with chronic cough, it shouldnot be assumed that GERD has been defini-tively ruled out as a cause of cough simply




because there is a history of antireflux surgery.Level of evidence, low; benefit, substantial; grade ofrecommendation, B

5. In patients with chronic cough, while teststhat link GERD with cough suggest a potentialcause-effect relationship, a definitive diagnosisof cough due to GERD requires that coughnearly or completely disappear with antirefluxtreatment. Level of evidence, low; benefit, substan-tial; grade of recommendation, B

6. In patients with chronic cough being eval-uated for GERD, the 24-h esophageal pH-mon-itoring test is the most sensitive and specifictest; however, it is recommended that the testresults be interpreted as normal only whenconventional indexes for acid reflux are withinthe normal range and no reflux-induced coughsappear during the monitoring study. Level ofevidence, low; benefit, substantial; grade of recom-mendation, B

7. In patients with cough who are undergo-ing 24-h monitoring, a low percentage ofcoughs associated with (or induced by) refluxdoes not exclude a diagnosis of cough due toGERD. Level of evidence, low; benefit, substantial;grade of recommendation, B

8. In patients with cough due to GERD, thedegree of abnormality noted in the esophagealpH-monitoring variables, such as the frequencyand duration of reflux events, does not directlycorrelate with the severity of the patients’cough. Level of evidence, low; benefit, substantial;grade of recommendation, B

9. In diagnosing nonacid GERD as the causeof cough, barium esophagography may be theonly available test to reveal GER of potentialpathologic significance in this setting (see thediscussion regarding esophageal impedancemonitoring in the “Laboratory Testing” sec-tion). When this is the case, barium esophagog-raphy is the test of choice to reveal GER ofpotential pathologic significance. Level of evi-dence, low; benefit, substantial; grade of recommen-dation, B

10. In patients with cough due to GERD,normal esophagoscopy findings do not rule outGERD as the cause of cough. Level of evidence,low; benefit, substantial; grade of recommendation,B

11. For patients fitting the clinical profile forcough due to GERD, it is recommended thattreatment be initially started in lieu of testing.Level of evidence, low; benefit, substantial; grade ofrecommendation, B

12. For patients fitting the clinical profile forcough due to GERD, the performance of 24-h

esophageal pH monitoring is recommended ontherapy when cough does not improve or re-solve to assist in determining whether the ther-apy needs to be intensified or if medical therapyhas failed. Level of evidence, low; benefit, substan-tial; grade of recommendation, B

13. For patients with chronic cough, the fol-lowing tests are not routinely recommended tolink cough with GERD: (a) assessing for lipid-laden macrophages in BAL fluid and inducedsputum, because this test has not been studiedin patients with chronic cough and because apositive test result is not specific for aspiration;(b) exhaled nitric oxide measurements, becausethey do not appear to be helpful in diagnosingcough due to GERD; (c) a Bernstein test, be-cause a negative Bernstein test result cannot beused to exclude the diagnosis of cough due toGERD; and (d) inhaled tussigenic challengeswith capsaicin, because they are not specific forcoughs due to GERD and because the testresult can be positive in patients with GERDwithout cough. Level of evidence, low; benefit,conflicting; grade of recommendation, I

Treatment

Based on the apparent heterogeneity of patientpopulations with differing pathogenetic mechanismsand differing risk factors that can adversely affectGERD, it is not likely that all patients will theoreti-cally respond to the same treatment. A review of theliterature on the treatment of cough due to GERD,which is summarized in Table 2, supports this state-ment. The review suggests the following: that whenmedical therapy is effective, some patients withcough due to GERD will favorably respond to acidsuppression therapy alone69,71; that proton pumpinhibition may be effective when H2-antagonism hasbeen ineffective57; that prokinetic therapy and diet,when added to proton pump inhibition, may beeffective when proton pump inhibition alone hasbeen ineffective57; that while medial therapy mayeliminate cough within 8 weeks in some patients,71 itmay still lead to a favorable outcome despite takingmonths before it starts to work9,53; and that cough inother patients will only improve or be eliminatedwith antireflux surgery, after cough in these patientshas failed to improve with maximal medical therapythat includes an intensive antireflux diet, maximumacid suppression, and prokinetic therapy.24 Thesefindings have come from prospective and retrospec-tive before-and-after intervention trials, and fromtwo imperfectly designed (see Table 2), prospective,




Table 2—Evidence for the Treatment of Chronic Cough Due to GERD*

Treatment Dosing Study IDPatients,

No. Age Results

Diet � LS, antacids,cimetidine

qid300 mg hs

Prospective, before/afterintervention, uncontrolled,unblinded15

5 67.2 yr Cough cured 100%

Diet � LS,metoclopramide and/or H2-blockers

Not statedNot stated


9 52 � 11 yr Cough cured 100%; time to cure, 161 � 74 d

Diet � LS,metoclopramide and/or H2-blockers



28 56 � 12 yr Cough cured 100%; time to cure, 179 � 205 d

Diet � LS, antacids,metoclopramide,cimetidine, surgery

Not stated,10 mg tid300, mg qid

Retrospective, before/afterintervention, uncontrolled,unblinded82

20 Adult Cough cured 90%; 70% cured with medication in 3mo, 20% with surgery

Diet � LS, omeprazoleor surgery

20–40 mg/d Prospective, before/afterintervention, uncontrolled,unblinded83

25 Adult Cough or hoarseness improved in 80%; only patientswithout heartburn had full resolution of cough

Cisapride,†domperidone

0.2 mg/kg tid0.2 mg/kg tid


55 Infants Cough improved 64.5%; no significant differencebetween drugs

Domperidone 2 mg/kg/ddivided qid


18 1 mo to 12.7 yr Cough improved 100% (p � 0.05)

Diet Prokinetic and/orH2-blockers



20 58 � 17 yr Cough cured in 97%; productive coughs in all patients

Surgery Prospective, before/afterintervention, uncontrolled,unblinded86

17 Adult 100% cure if esophageal motility was normal


5 Adult 100% cured in highly selected patients who werethought to be aspirators; only 5/100 patients weresuitable


13 Adult Cough cured in 84.6%; only 13/140 patientsconsidered suitable

Diet, prokinetics and/orPPIs

Cisapride or metoclopra-pramide 10 mg qid,omeprazole up to 40mg and lansoprazole,30 mg bid

Retrospective, before/afterintervention, uncontrolled,unblinded57

56 57 yr Cough cured or improved in 86%; 17/56 had notresponded to H2-blockers; 5/56 required up to 8weeks to improve, and 1/55 up to 12 weeks; 24/56only required PPIs, 18/56 required PPIs plusprokinetics

Omeprazole Omeprazole, 40 mg/d, orplacebo for 8 wk

Prospective, double-blind,randomized, placebo-controlled, cross-overtrial71

29 Adults Cough improved (p � 0.02) with omeprazolecompared to placebo; proper statistics were notutilized to analyze crossover design

Omeprazole Omeprazole, 40 mg bid,or placebo for 12 wk

Prospective, double-blind,randomized, placebo-controlled trial69

23 Adults Of 17 patients with abnormal findings of 24-hesophageal pH monitoring studies, 6 significantlyimproved after drug therapy compared to placebo; itis not known whether the remaining patients alsohad GERD because no further testing wasperformed and treatment was not intensified


119 Adults Of 119 patients with inadequate control of respiratorysymptoms with medical therapy, 42 had chroniccough as chief complaint; surgery cured orimproved cough in 85% at 6 mo after surgery

Surgery Prospective before/afterintervention, uncontrolled,unblinded89

35 Adults Of 35 patients with chronic cough, cough improved orwas cured in 56% at 12 mo; patients did not appearto have been selected based upon lack of responseto medical treatment


21 Adults At 1 yr following surgery, cough had significantlyimproved (p � 0.01); improvement was reported by86% of patients, with complete resolution in 62%;prior to surgery, cough had failed to improve withintensive medical therapy


8 54.2 � 10.8 yr At 1 yr following surgery, cough had significantlyimproved (p � 0.001); all patients reportedimprovement in cough; prior to surgery, patientshad failed to respond to maximal medical therapyeven though there was total/near-total elimination ofesophageal acid

*LS � lifestyle changes (such as not wearing constricting clothing and head of bed elevation). Diets were not well-described in any study and werenot consistently described between studies.

†Cisapride is off the market in the United States but is available in other countries.




double-blind, randomized placebo-controlled tri-als69,71 that have compared the effects of omeprazolewith placebo.

While the optimal way to treat all patients withcough due to GERD has yet to be determined,medical therapy in adults, based on the results ofprospective and retrospective before-and-after inter-vention trials, when not limited to acid suppressiontherapy alone, has improved cough in 70 to 100% ofpatients. When cough in adults has not improvedwith medical therapy, surgery has improved cough 6to 12 months following surgery in prospective, be-fore-and-after intervention trials33,67 in approxi-mately 85% of patients.

Until multiple randomized controlled clinical trialsprovide guidance on which treatment regimens willbe successful in which groups of subjects, the panelby consensus recommends that clinicians must beaware of the spectrum of treatment options that mayfavorably affect GERD (Table 3) and that an inten-sive medical treatment regimen that includes allmedical therapeutic options listed in Table 3 beinstituted before considering antireflux surgery.17

While there has been an anecdotal report72 suggest-ing that one patient may have developed an intrac-table cough due to receiving omeprazole therapy,the patient was not rechallenged with the drug tosolidify the association. Nevertheless, if drug-in-duced cough is considered to be a clinical possibility,the agent should be discontinued to determinewhether it might be playing a role before recom-mending surgery.

An intensive medical treatment regimen includesthe following: (1) an antireflux diet that includes no� 45 g of fat in 24 h and no coffee, tea, soda,chocolate, mints, citrus products, including toma-

toes37,73 or alcohol,41 no smoking,37 and limitingvigorous exercise that will increase intraabdominalpressure39,40; (2) acid suppression with a protonpump inhibitor (PPI); and (3) prokinetic therapy.Even though nocturnal breakthrough gastric acidproduction has been noted in healthy control sub-jects and patients with GERD who are receivingPPIs, the addition of a nocturnal H2-antagonist hasnot been shown to improve clinical outcomes.66,74,75

Because comorbid diseases such as obstructive sleepapnea51,76 or therapy for comorbid conditions (eg,nitrates, progesterone, and calcium channel block-ers)37,77 can potentially make GERD more difficultto control, routinely consider trying to mitigate theinfluences of these factors whenever possible. Be-cause coughing can induce GER events, it should beappreciated that less than optimal results may occurdue to a cough-GER self-perpetuating cycle unlessall causes of cough are adequately treated.28 Reli-ance on acid suppression alone may not only fail toadequately control the cough but also potentiallyplace the patient at increased risk of community-acquired pneumonia. According to a retrospectivecase-control analysis, which was nested in a cohort ofincident users of acid-suppressive drugs for at least 1year,78 the adjusted risk for pneumonia among per-sons currently using PPIs compared with those whostopped receiving the PPIs was 1.89. Current usersof histamine-2 receptor antagonists had a 1.63-foldincreased risk of pneumonia compared with thosepatients who had stopped using them. These datatranslated into approximately one case of pneumoniafor every 100 years of patient exposure.79 A prospec-tive, randomized, controlled clinical trial will benecessary to confirm the accuracy of these results.

While the role of antireflux surgery has not beenclearly defined, it appears that antireflux surgery hasbeen able to effectively eliminate or significantlyimprove cough that is unresponsive to intensiveantireflux medical therapy. The panel by consensusrecommends that antireflux surgery be considered inpatients who meet the following criteria: (1) thefindings of a 24-h esophageal pH-monitoring studybefore treatment are positive, as defined above; (2)patients fit the clinical profile suggesting that GERDis the likely cause of their cough (Table 1); (3) coughhas not improved after a minimum of 3 months ofintensive therapy (Table 3), and serial esophagealpH-monitoring studies or other objective studies (eg,barium esophagography, esophagoscopy, or gastric-emptying study with solids) performed on therapyshow that intensive medical therapy has failed tocontrol the reflux disease and GERD is still the likelycause of cough; and (4) patients express the opinionthat their persisting cough does not allow them asatisfactory quality of life. While one small prospec-

Table 3—Spectrum of Options for Treating ChronicCough Due to GERD

Treatment Options

Antireflux medical therapy DietLifestyle changes

SmokingExercisingConsuming alcohol

MedicationsAcid suppressionProkinetic

Address risk factorsTreat other causes of coughTreat comorbid conditions

ObesityObstructive sleep apnea

Consider changing medications forcomorbid conditions

Antireflux surgery




tive, uncontrolled and unblinded, descriptive study80

suggested that an assessment of gastric emptyingwould not assist in patient management, anothersmall, prospective, before-and-after intervention tri-al24 found that a gastric-emptying study with solidsprovided the only insight as to why prior antirefluxsurgery had failed to control cough. Redo antirefluxsurgery that included a gastric-emptying procedurein three of four patients was successful.24 While avariety of endoscopically assisted interventional tech-niques have been developed and performed in pa-tients with classic GI GERD symptoms,81 theseprocedures should be considered experimental atthis time in the population of patients with cough.

Recommendations

14. In patients who meet the clinical profilepredicting that silent GERD is the likely causeof chronic cough or in patients with chroniccough who also have prominent upper GI symp-toms consistent with GERD, an empiric trial ofmedical antireflux therapy is recommended.Level of evidence, low; benefit, substantial; grade ofrecommendation, B

15. For treating the majority of patients withchronic cough due to GERD, the followingmedical therapies are recommended: (a) di-etary and lifestyle modifications; (b) acid sup-pression therapy; and (c) the addition of proki-netic therapy either initially or if there is noresponse to the first two therapies. The re-sponse to these therapies should be assessedwithin 1 to 3 months. Level of evidence, expertopinion; benefit, substantial; grade of recommenda-tion, E/A

16. In patients in which this empiric treat-ment fails, it cannot be assumed that GERD hasbeen ruled out as a cause of chronic cough;rather, the objective investigation for GERD isthen recommended because the empiric ther-apy may not have been intensive enough ormedical therapy may have failed. Level of evi-dence, expert opinion; benefit, substantial; grade ofrecommendation, E/A

17. In some patients, cough due to GERDwill favorably respond to acid suppression ther-apy alone; proton pump inhibition may be ef-fective when H2-antagonism has been ineffec-tive; prokinetic therapy and diet, when added toproton pump inhibition, may be effective whenproton pump inhibition alone has been ineffec-tive. Level of evidence, low; benefit, substantial;grade of recommendation, B

18. Patients requiring an intensive medical

treatment regimen should be treated with thefollowing: (a) antireflux diet that includes no> 45 g of fat in 24 h and no coffee, tea, soda,chocolate, mints, citrus products, including to-matoes, or alcohol, no smoking, and limitingvigorous exercise that will increase intra-ab-dominal pressure; (b) acid suppression with aPPI; (c) prokinetic therapy; and (d) efforts tomitigate the influences of comorbid diseasessuch as obstructive sleep apnea or therapy forcomorbid conditions (eg, nitrates, progester-one, and calcium channel blockers) wheneverpossible. Level of evidence, expert opinion; benefit,substantial; grade of recommendation, E/A

19. In patients with chronic cough due toGERD that has failed to improve with the mostmaximal medical therapy, which includes anintensive antireflux diet and lifestyle modifica-tion, maximum acid suppression, and prokinetictherapy, and the rest of the spectrum of treat-ment options in Table 3, cough may only im-prove or be eliminated with antireflux surgery.Level of evidence, low; benefit, substantial; grade ofrecommendation, B

20. In patients who meet the following crite-ria, antireflux surgery is the recommendedtreatment: (a) findings of a 24-h esophagealpH-monitoring study before treatment is posi-tive, as defined above; (b) patients fit the clini-cal profile suggesting that GERD is the likelycause of their cough (Table 1); (c) cough has notimproved after a minimum of 3 months ofintensive therapy (Table 3), and serial esopha-geal pH-monitoring studies or other objectivestudies (eg, barium esophagography, esopha-goscopy, and gastric-emptying study with sol-ids) performed while the patient receives ther-apy show that intensive medical therapy hasfailed to control the reflux disease and thatGERD is still the likely cause of cough; and (d)patients express the opinion that their persist-ing cough does not allow them a satisfactoryquality of life. Level of evidence, expert opinion;benefit, substantial; grade of recommendation, E/A

Summary of Recommendations

1. In patients with chronic cough due toGERD, the term acid reflux disease, unlessit can be definitively shown to apply, shouldbe replaced by the more general term refluxdisease so as not to mislead the cliniciansinto thinking that all patients with coughdue to GERD should improve with acid-




suppression therapy. Level of evidence, expertopinion; benefit, substantial; grade of recom-mendation, E/A

2. In patients with chronic cough whoalso complain of typical and frequent GIcomplaints such as daily heartburn and re-gurgitation, especially when the findings ofchest-imaging studies and/or clinical syn-drome are consistent with an aspirationsyndrome, the diagnostic evaluation shouldalways include GERD as a possible cause.Level of evidence, low; benefit, substantial;grade of recommendation, B

3. Patients with chronic cough who haveGI symptoms that are consistent withGERD or who fit the clinical profile de-scribed in Table 1, should be considered tohave a high likelihood of having GERD andshould be prescribed antireflux treatmenteven when they have no GI symptoms. Levelof evidence, low; benefit, substantial; grade ofrecommendation, B

4. In patients with chronic cough, itshould not be assumed that GERD has beendefinitively ruled out as a cause of coughsimply because there is a history of antire-flux surgery. Level of evidence, low; benefit,substantial; grade of recommendation, B

5. In patients with chronic cough, whiletests that link GERD with cough suggest apotential cause-effect relationship, a defin-itive diagnosis of cough due to GERD re-quires that cough nearly or completely dis-appear with antireflux treatment. Level ofevidence, low; benefit, substantial; grade of rec-ommendation, B

6. In patients with chronic cough beingevaluated for GERD, the 24-h esophagealpH-monitoring test is the most sensitive andspecific test; however, it is recommendedthat the test results be interpreted as nor-mal only when conventional indexes for acidreflux are within the normal range and noreflux-induced coughs appear during themonitoring study. Level of evidence, low; ben-efit, substantial; grade of recommendation, B

7. In patients with cough who are under-going 24-h monitoring, a low percentage ofcoughs associated with (or induced by) re-flux does not exclude a diagnosis of coughdue to GERD. Level of evidence, low; benefit,substantial; grade of recommendation, B

8. In patients with cough due to GERD,the degree of abnormality noted in the

esophageal pH-monitoring variables, suchas the frequency and duration of refluxevents, does not directly correlate with theseverity of the patients’ cough. Level ofevidence, low; benefit, substantial; grade of rec-ommendation, B

9. In diagnosing nonacid GERD as thecause of cough, barium esophagographymay be the only available test to reveal GERof potential pathologic significance in thissetting (see the discussion regarding esoph-ageal impedance monitoring in the “Labo-ratory Testing” section). When this is thecase, barium esophagography is the test ofchoice to reveal GER of potential patho-logic significance. Level of evidence, low; ben-efit, substantial; grade of recommendation, B

10. In patients with cough due to GERD,normal esophagoscopy findings do not ruleout GERD as the cause of cough. Level ofevidence, low; benefit, substantial; grade of rec-ommendation, B

11. For patients fitting the clinical profilefor cough due to GERD, it is recommendedthat treatment be initially started in lieu oftesting. Level of evidence, low; benefit, sub-stantial; grade of recommendation, B

12. For patients fitting the clinical profilefor cough due to GERD, the performance of24-h esophageal pH monitoring is recom-mended on therapy when cough does notimprove or resolve to assist in determiningwhether the therapy needs to be intensifiedor if medical therapy has failed. Level ofevidence, low; benefit, substantial; grade of rec-ommendation, B

13. For patients with chronic cough, thefollowing tests are not routinely recom-mended to link cough with GERD: (a) as-sessing for lipid-laden macrophages in BALfluid and induced sputum, because this testhas not been studied in patients withchronic cough and because a positive testresult is not specific for aspiration; (b) ex-haled nitric oxide measurements, becausethey do not appear to be helpful in diagnos-ing cough due to GERD; (c) a Bernsteintest, because a negative Bernstein test re-sult cannot be used to exclude the diagnosisof cough due to GERD; and (d) inhaledtussigenic challenges with capsaicin, be-cause they are not specific for coughs due toGERD and because the test result can bepositive in patients with GERD without




cough. Level of evidence, low; benefit, conflict-ing; grade of recommendation, I

14. In patients who meet the clinical pro-file predicting that silent GERD is the likelycause of chronic cough or in patients withchronic cough who also have prominentupper GI symptoms consistent with GERD,an empiric trial of medical antireflux ther-apy is recommended. Level of evidence, low;benefit, substantial; grade of recommendation, B

15. For treating the majority of patientswith chronic cough due to GERD, the fol-lowing medical therapies are recommend-ed: (a) dietary and lifestyle modifications;(b) acid suppression therapy; and (c) theaddition of prokinetic therapy either ini-tially or if there is no response to the firsttwo therapies. The response to these thera-pies should be assessed within 1 to 3months. Level of evidence, expert opinion; ben-efit, substantial; grade of recommendation, E/A

16. In patients in which this empirictreatment fails, it cannot be assumed thatGERD has been ruled out as a cause ofchronic cough; rather, the objective investi-gation for GERD is then recommended be-cause the empiric therapy may not havebeen intensive enough or medical therapymay have failed. Level of evidence, expertopinion; benefit, substantial; grade of recom-mendation, E/A

17. In some patients, cough due to GERDwill favorably respond to acid suppressiontherapy alone; proton pump inhibition maybe effective when H2-antagonism has beenineffective; prokinetic therapy and diet,when added to proton pump inhibition, maybe effective when proton pump inhibitionalone has been ineffective. Level of evidence,low; benefit, substantial; grade of recommenda-tion, B

18. Patients requiring an intensive medi-cal treatment regimen should be treatedwith the following: (a) antireflux diet thatincludes no > 45 g of fat in 24 h and nocoffee, tea, soda, chocolate, mints, citrusproducts, including tomatoes, or alcohol, nosmoking, and limiting vigorous exercise thatwill increase intraabdominal pressure; (b)acid suppression with a PPI; (c) prokinetictherapy; and (d) efforts to mitigate the in-fluences of comorbid diseases such as ob-structive sleep apnea or therapy for comor-bid conditions (eg, nitrates, progesterone,

and calcium channel blockers) wheneverpossible. Level of evidence, expert opinion;benefit, substantial; grade of recommendation,E/A

19. In patients with chronic cough due toGERD that has failed to improve with themost maximal medical therapy, which in-cludes an intensive antireflux diet and life-style modification, maximum acid suppres-sion, and prokinetic therapy, and the rest ofthe spectrum of treatment options in Table3, cough may only improve or be eliminatedwith antireflux surgery. Level of evidence,low; benefit, substantial; grade of recommenda-tion, B

20. In patients who meet the followingcriteria, antireflux surgery is the recom-mended treatment: (a) findings of a 24-hesophageal pH-monitoring study beforetreatment is positive, as defined above; (b)patients fit the clinical profile suggestingthat GERD is the likely cause of their cough(Table 1); (c) cough has not improved after aminimum of 3 months of intensive therapy(Table 3), and serial esophageal pH-moni-toring studies or other objective studies (eg,barium esophagography, esophagoscopy,and gastric-emptying study with solids) per-formed while the patient receives therapyshow that intensive medical therapy hasfailed to control the reflux disease and thatGERD is still the likely cause of cough; and(d) patients express the opinion that theirpersisting cough does not allow them asatisfactory quality of life. Level of evidence,expert opinion; benefit, substantial; grade ofrecommendation, E/A

References1 Koufman J. Larynogopharyngeal reflux is different from

classic gastroesophageal reflux disease. Ear Nose Throat J2002; 81(suppl):7–9

2 Koufman J. The otolaryngologic manifestations of gastro-esophageal reflux disease (GERD: a clinical investigation of225 patients using ambulatory 24-hour pH monitoring and anexperimental investigation of the role of acid and pepsin inthe development of laryngeal injury. Laryngoscope 1991;101(suppl):1–78

3 Locke G III, Talley NJ, Fett SL, et al. Prevalence and clinicalspectrum of gastroesophageal reflux: a population-basedstudy in Olmsted County, Minnesota. Gastroenterology 1997;112:1448–1456

4 Irwin R, French CL, Curley FJ, et al. Chronic cough due togastroesophageal reflux: clinical, diagnostic, and pathogenetic




aspects. Chest 1993; 104:1511–15175 Ossakow S, Elta G, Colturi T, et al. Esophageal reflux and

dysmotility as the basis for persistent cervical symptoms. AnnOtol Rhinol Laryngol 1987; 96:387–392

6 Richter J. Extraesophageal presentations of gastroesophagealreflux disease: an overview. Am J Gastroenterol 2000;95(suppl):S1–S3

7 Shaheen N, Ransohoff DF. Gastroesophageal reflux, Barrettesophagus, and esophageal cancer: scientific review. JAMA2002; 287:1972–1981

8 Inadomi J, Sampliner R, Lagergren J, et al. Screening andsurveillance for Barrett esophagus in high-risk groups: acost-utility analysis. Ann Intern Med 2003; 138:176–186

9 Irwin R, Curley FJ, French CL. Chronic cough: the spectrumand frequency of causes, key components of the diagnosticevaluation, and outcome of specific therapy. Am Rev RespirDis 1990; 141:640–647

10 Balgos A, Reyes-Pingol R, Siasoco BR, et al. Chronic persis-tent cough: local expertise in diagnosis using the anatomicand diagnotic protocol, spectrum and frequency of causes andoutcome of specific treatment. Phil J Chest Dis 1993; 1:6–11

11 Marchesani F, Cecarini L, Pela R, et al. Causes of chronicpersistent cough in adult patients: the results of systematicmanagement protocol. Monaldi Arch Chest Dis 1998; 53:510–514

12 McGarvey L, Heaney LG, Lawson JT, et al. Evaluation andoutcome of patients with chronic non-productive cough usinga comprehensive diagnostic protocol. Thorax 1998; 53:738–743

13 Brightling C, Ward R, Goh KL, et al. Eosinophilic bronchitisis an important cause of chronic cough. Am J Respir Crit CareMed 1999; 160:406–410

14 Palombini B, Villanova CAC, Araujo E, et al. A pathogenictriad in chronic cough: asthma, postnasal drip syndrome, andgastroesophageal reflux disease. Chest 1999; 116:279–284

15 Irwin R, Corrao WM, Pratter MR. Chronic persistent coughin the adult: the spectrum and frequency of causes andsuccessful outcome of specific therapy. Am Rev Respir Dis1981; 123:413–417

16 French C, Irwin RS, Curley FJ, et al. Impact of chronic coughon quality of life. Arch Intern Med 1998; 158:1657–1661

17 Irwin R, Madison JM. Diagnosis and treatment of chroniccough due to gastro-esophageal reflux disease and postnasaldrip syndrome. Pulm Pharmacol Ther 2002; 15:261–266

18 Ing A Jr, Ngu MC, Breslin ABX. Pathogenesis of chronicpersistent cough associated with gastroesophageal reflux.Am J Respir Crit Care Med 1994; 149:160–167

19 Irwin R, Madison JM, Fraire AE. The cough reflex and itsrelation to gastroesophageal reflux. Am J Med 2000;108(suppl):73S–78S

20 Wong R, Hanson DG, Waring PJ, et al. ENT Manifestationsof gastroesophageal reflux. Am J Gastroenterol 2000; 95:S15–S22

21 Wolfe J, Bone RC, Ruth WE. Diagnosis of gastric aspirationby fiberoptic bronchoscopy. Chest 1976; 70:458–459

22 Markowitz D, Irwin, RS. Aspiration. In: Irwin R, Rippe JM,eds. Intensive care medicine. Philadelphia, PA: LippincottWilliams & Wilkins, 2003; 591–597

23 Irwin R, French, CT. Cough and gastroesophageal reflux:identifying cough and assessing the efficacy of cough-modi-fying agents. Am J Med 2001; 111(suppl):455–505

24 Irwin R, Zawacki JK, Wilson MM, et al. Chronic cough due togastroesophageal reflux disease: failure to resolve despitetotal/near-total elimination of esophageal acid. Chest 2002;121:1132–1140

25 Knight R, Wells JR, Parrish RS. Esophageal dysmotility as animportant co-factor in extraesophageal manifestations of gas-

troesophageal reflux. Laryngoscope 2000; 110:1462–146626 Fouad Y, Katz PO, Hatlebakk JG, et al. Ineffective esopha-

geal motility: the most common motility abnormality inpatients with GERD-associated respiratory symptoms. Am JGastroenterol 1999; 94:1464–1467

27 Patterson W, Murat BW. Combined ambulatory esophagealmanometry and dual-probe pH-metry in evaluation of pa-tients with chronic unexplained cough. Dig Dis Sci 1994;39:1117–1125

28 Irwin R, Boulet L-P, Cloutier MM, et al. Managing cough asa defense mechanism and as a symptom: a consensus panelreport of the American College of Chest Physicians. Chest1998; 114(suppl):133S–181S

29 Ferrari M, Olivieri M, Sembenini C, et al. Tussive effect ofcapsaicin in patients with gastroesophageal reflux withoutcough. Am J Respir Crit Care Med 1995; 151:557–561

30 O’Connell F, Thomas VE, Pride NB, et al. Capsaicin coughsensitivity decreases with successful treatment of chroniccough. Am J Respir Crit Care Med 1994; 150:374–380

31 Mello C, Irwin RS, Curley FJ. Predictive values of thecharacter, timing, and complications of chronic cough indiagnosing its cause. Arch Intern Med 1996; 156:997–1003

32 Smyrnios N, Irwin RS, Curley FJ. Chronic cough with ahistory of excessive sputum production. Chest 1995; 108:991–997

33 Novitsky Y, Zawacki JK, Irwin RS, et al. Chronic cough due togastroesophageal reflux disease: efficacy of antireflux surgery.Surg Endosc 2002; 16:567–571

34 de Groen P, Lubbe DF, Hirsch LJ, et al. Esophagitisassociated with the use of alendronate. N Engl J Med 1996;335:1016–1021

35 Lazenby J, Guzzo MR, Harding SM, et al. Oral corticoste-roids increase esophageal acid contact times in patients withstable asthma. Chest 2002; 121:625–634

36 Crowell M, Zayat EN, Lacy BE, et al. The effects of aninhaled �2-adrenergic agonist on lower esophageal function: adose-response study. Chest 2001; 120:1184–1189

37 Richter J, Castell DO. Drugs, foods, and other substances inthe cause and treatment of reflux esophagitis. Med ClinNorth Am 1981; 65:1223–1234

38 Nilsson M, Johnsen R, Ye W, et al. Obesity and estrogen asrisk factors for gastroesophageal reflux symptoms. JAMA2003; 290:66–72

39 Clark C, Kraus BB, Sinclair J, et al. Gastroesophageal refluxinduced by exercise in healthy volunteers. JAMA 1989;261:3599–3601

40 Kraus B, Sinclair JW, Castell DO. Gastroesophageal reflux inrunners: characteristics and treatment. Ann Intern Med 1990;112:429–433

41 Vitale G, Cheadle WG, Patel B, et al. The effect of alcohol onnocturnal gastroesophageal reflux. JAMA 1987; 258:2077–2079

42 Nagler R, Spiro HM. Persistent gastroesophageal refluxinduced during prolonged gastric intubation. N Engl J Med1963; 269:495–500

43 Vinnik I, Kern F Jr. The effect of gastric intubation onesophageal pH. Gastroenterology 1964; 47:388–394

44 Young L, Hadjiliadis D, Davis RD, et al. Lung transplantationexacerbates gastroesophageal reflux disease. Chest 2003; 124:1689–1693

45 Kopec S, Irwin RS, Umali-Torres CB, et al. The postpneu-monectomy state. Chest 1998; 114:1158–1184

46 Min F, Tarlo SM, Bargman J, et al. Prevalence and causes ofcough in chronic dialysis patients: a comparison betweenhemodialysis and peritoneal dialysis patients. Adv Perit Dial2000; 16:129–133

47 Twardowski Z, Khanna R, Nolph KD, et al. Intra-abdominal




pressures during natural activities in patients treated withcontinuous ambulatory peritoneal dialysis. Nephron 1986;44:129–135

48 Harding S, Richter JE. The role of gastroesophageal reflux inchronic cough and asthma. Chest 1997; 111:1389–1402

49 Ing A, Ngu MC, Breslin ABX. Obstructive sleep apnea andgastroesophageal reflux. Am J Med 2000; 108:120S–125S

50 Valipour A, Makker HK, Hardy R, et al. Symptomaticgastroesophageal reflux in subjects with a breathing sleepdisorder. Chest 2002; 121:1748–1753

51 Green B, Broughton WA, O’Connor JB. Marked improve-ment in nocturnal gastroesophageal reflux in a large cohort ofpatients with obstructive sleep apnea treated with continuouspositive airway pressure. Arch Intern Med 2003; 163:41–45

52 Hu F, Preston RA, Post JC, et al. Mapping of a gene forsevere pediatric gastroesophageal reflux to chromosome13q14. JAMA 2000; 284:325–334

53 Irwin R, Zawacki JK, Curley FJ, et al. Chronic cough as thesole presenting manifestation of gastroesophageal reflux. AmRev Respir Dis 1989; 140:294–300

54 Ing A, Ngu MC. Cough and gastro-oesophageal reflux.Lancet 1999; 353:944–946

55 Pandolfino J, Richter JE, Ours T, et al. Ambulatory esopha-geal pH monitoring using a wireless system. Am J Gastroen-terol 2003; 98:740–749

56 Vaezi M, Richter JE. Twenty-four-hour ambulatory esopha-geal pH monitoring in the diagnosis of acid reflux-relatedchronic cough. South Med J 1997; 90:305–311

57 Poe R, Kallay MC. Chronic cough and gastroesophagealreflux disease: experience with specific therapy for diagnosisand treatment. Chest 2003; 123:679–684

58 Vela M, Camacho-Lobato L, Srinivasan R, et al. Simultaneousintraesophageal impedance and pH measurement of acid andnonacid gastroesophageal reflux: effect of omeprazole. Gas-troenterology 2001; 120:1599–1606

59 Sifram D, Holloway R, Silney J, et al. Acid, nonacid, and gasreflux in patients with gastroesophageal dysmotility as a causeof chronic persistent cough. Gastroenterology 2001; 120:1588–1598

60 Kastelik J, Redinton AE, Aziz I, et al. Abnormal oesophagealmotility in patients with chronic cough. Thorax 2003; 58:699–702

61 Belafsky P, Postma GN, Koufman JA. The validity andreliability of the reflux finding score (FRS). Laryngoscope2001; 111:1313–1317

62 Corwin R, Irwin RS. The Lipid-laden alveolar macrophage asa marker of aspiration in parenchymal lung disease. Am RevRespir Dis 1985; 132:576–581

63 Parameswaran K, Anvari M, Efthimiadis A, et al. Lipid-ladenmacrophages in induced sputum are a marker of oropharyn-geal reflux and possible gastric aspiration. Eur Respir J 2000;16:1119–1122

64 Chatkin J, Ansarin K, Silkoff PE, et al. Exhaled nitric oxide asa noninvasive assessment of chronic cough. Am J Respir CritCare Med 1999; 159:1810–1813

65 Parameswaran K, Allen CJ, Kamada D, et al. Sputum cellcounts and exhaled nitric oxide in patients with gastroesoph-ageal reflux, and cough or asthma. Can Respir J 2001;8:239–244

66 Peghini P, Katz PO, Bracy NA, et al. Nocturnal recovery ofgastric acid secretion with twice-daily dosing of proton pumpinhibitors. Am J Gastroenterol 1998; 93:763–767

67 Allen C, Anvari M. Gastro-oesophageal reflux related coughand its response to laparoscopic fundoplication. Thorax 1998;53:963–968

68 Numans M, Lau J, de Wit NJ, et al. Short-term treatmentwith proton-pump inhibitors as a test for gastroesophageal

reflux disease: a meta-analysis of diagnostic test characteris-tics. Ann Intern Med 2004; 140:518–527

69 Ours T, Kavuru MS, Schilz RJ, et al. A prospective evaluationof esophageal testing and a double-blind, randomized study ofomeprazole in a diagnostic and therapeutic algorithm forchronic cough. Am J Gastroenterol 1999; 94:3131–3138

70 Irwin R, Zawacki JK. Accurately diagnosing and successfullytreating chronic cough due to gastroesophageal reflux diseasecan be difficult. Am J Gastroenterol 1999; 94:3095–3098

71 Kiljander T, Salomaa ERM, Hietanen EK, et al. Chronic coughand gastro-oesophageal reflux: a double-blind placebo-con-trolled study with omeprazole. Eur Respir J 2000; 16:633–638

72 Howaizi MDC. Omeprazole-induced intractable cough. AnnPharmacother 2003; 37:1607–1609

73 Irwin R, Madison JM. The diagnosis and treatment of cough.N Engl J Med 2000; 343:1715–1721

74 Fackler W, Ours TM, Vaezi MF, et al. Long-term effect of H2RA therapy on nocturnal gastric acid breakthrough. Gastro-enterology 2002; 122:625–632

75 Orr W, Harnish MJ. The efficacy of omeprazole twice dailywith supplemental H2 blockade at bedtime in the suppressionof nocturnal oesophageal and gastric acidity. Aliment Phar-macol Ther 2003; 17:1553–1558

76 Kerr P, Shoenut JP, Millar T, et al. Nasal CPAP reducesgastroesophageal reflux in obstructive sleep apnea syndrome.Chest 1992; 101:1539–1544

77 Kitchin L, Castell DO. Rationale and efficacy of conservativetherapy for gastroesophageal reflux disease. Arch Intern Med1991; 151:448–454

78 Laheij R, Sturkenboom, MCJM, Hassing R-J, et al. Risk ofcommunity-acquired pneumonia and use of gastric acid-suppressive drugs. JAMA 2004; 292:1955–1960

79 Gregor J. Acid suppression and pneumonia: a clinical indica-tion for rational prescribing. JAMA 2004; 292:2012–2013

80 Kastelik J, Jackson W, Davies TW, et al. Measurement ofgastric emptying in gastroesophageal reflux-related chroniccough. Chest 2002; 122:2038–2041

81 Triadafilopoulos G. Changes in GERD symptom scorescorrelate with improvement in esophageal acid exposure afterthe Stretta procedure. Surg Endosc 2004; 18:1038–1044

82 Fitzgerald J, Allen CJ, Craven MA, et al. Chronic cough andgastroesophageal reflux. Can Med Assoc J 1989; 140:520–524

83 Waring J, Lacayo L, Hutner J, et al. Chronic cough andhoarseness in patients with severe gastroesophageal refluxdisease: diagnosis and response to therapy. Dig Dis Sci 1995;40:1093–1097

84 Dordal M, Baltazar MA, Roca I, et al. Nocturnal spasmodiccough in the infant: evolution after antireflux treatment.Allerg Immunol (Paris) 1994; 26:53–58

85 DuPont C, Molkhou P, Petrovic N, et al. Treatment usingmotilium of gastro-esophageal reflux associated with respiratorymanifestations in children. Ann Pediatr 1989; 36:148–150

86 DeMeester T, Bonavina L, Lascone C, et al. Chronic respi-ratory symptoms and occult gastroesophageal reflux: a pro-spective clinical trial and results of surgical therapy. Ann Surg1990; 211:337–345

87 Pellegrini C, DeMeester TR, Johnson LF, et al. Gastroesoph-ageal reflux and pulmonary aspiration: incidence, functionalabnormality, and results of surgical therapy. Surgery 1979;86:110–119

88 Giudicelli R, Dupin B, Surpas P, et al. Gastroesophageal refluxand respiratory manifestations: diagnostic approach, therapeuticindications and results. Ann Chir 1990; 47:552–554

89 So J, Zeitels SM, Rattner DW. Outcomes of atypical symp-toms attributed to gastroesophageal reflux treated by laparo-scopic fundoplication. Surgery 1998; 124:28–32




Chronic Cough Due to Gastroesophageal Reflux Disease...

Documents

Transcript of Chronic Cough Due to Gastroesophageal Reflux Disease...