Otolaryngology–Head and Neck Surgery (2010) 143, 765-771

ORIGINAL RESEARCH–LARYNGOLOGY AND NEUROLARYNGOLOGY

Correlation of pepsin-measured laryngopharyngeal

reflux disease with symptoms and signs

Lu Wang, MD, Xiong Liu, MD, You-li Liu, Fang-fang Zeng, Ting Wu, MS,

Chun-li Yang, Hai-yan Shen, MS, and Xiang-ping Li, MD, Guangzhou, China

No sponsorships or competing interests have been disclosed forthis article.

ABSTRACT

OBJECTIVE: Pepsin detection in throat sputum has been pos-ited as a reliable biological marker of laryngopharyngeal refluxdisease (LPRD). This study was designed to further correlatepepsin concentration with symptoms and signs of LPRD.STUDY DESIGN: Cross-sectional study.SETTING: Nanfang Hospital of Southern Medical University.SUBJECTS AND METHODS: Fifty-six laryngitis patientswere divided into a reflux laryngitis group and a chronic laryngitisgroup based on the reflux symptom index (RSI), reflux findingscore (RFS), and proton pump inhibitor treatment for two weeks.Oral and hypopharyngeal secretions from the study patients andfrom 15 healthy subjects were collected. Thirty-six obstructivesleep apnea (OSA) patients were divided into a mild-moderategroup and a severe group by the apnea-hypopnea index (AHI).Bedtime and first-morning oral secretions were collected. Enzyme-linked immunosorbent assay was used to measure the pepsinconcentration.RESULTS: In laryngitis patients, the total score of RSI and RFS(P � 0.05), and the symptoms, including clearing throat often,coughing, and sensing a lump in the throat (P � 0.006), were moresevere in the pepsin-positive group. No significant differenceswere found between the oral and hypopharyngeal secretions. InOSA, pepsin levels in the first-morning oral secretions were cor-related with AHI, mean SaO2, and mini SaO2 (P � 0.01). How-ever, RSIs were not significantly correlated with these indicators.CONCLUSION: Higher levels of pepsin in sputum were asso-ciated with higher RSI and RFS in cases of laryngitis. There wasno relationship between pepsin levels and RSI in cases of OSA.There were no differences of pepsin concentration in sputumcollection methods or in collection timing.

© 2010 American Academy of Otolaryngology–Head and NeckSurgery Foundation. All rights reserved.

Laryngopharyngeal reflux (LPR), the backflow of gastriccontents into the larynx, pharynx, trachea, and bron-

chus, contributes to several upper airway inflammatory dis-orders, which are referred to as laryngopharyngeal refluxdisease (LPRD). This term was adopted by the American

Received April 9, 2010; revised August 10, 2010; accepted August 13, 2010.

0194-5998/$36.00 © 2010 American Academy of Otolaryngology–Head and Necdoi:10.1016/j.otohns.2010.08.018

Academy of Otolaryngology–Head and Neck Surgery in its2002 position statement on LPR.1 LPR is associated withreflux laryngitis, obstructive sleep apnea (OSA), laryngealcancer, contact laryngeal ulcers, laryngomalacia, posteriorglottic edema and erythema, laryngeal granuloma, laryngo-spasm, stridor, subglottic stenosis, and otitis media witheffusion.2-4

To our knowledge, no method has been validated toperform a quantitative analysis of LPRD,5 which might bea main cause of many misdiagnosed LPR patients. Themethod suggested by Ford6 to combine a set of measure-ments to assess and manage LPRD has become increasinglyaccepted. Ford suggested that patients be first clinicallyevaluated with the reflux symptom index (RSI; history,symptoms � 13), the reflux finding score (RFS; videolaryn-goscopy � 7), and then be engaged in an empirical trial oflifestyle and dietary changes. After an initial assessment,clinicians should treat LPR for three months using protonpump inhibitors, the most potent form of acid suppressiontherapy available. Multi-channel impedance and pH moni-toring are to be used in only those patients whose symptomsare unchanged or have worsened after three months ofproton pump inhibitor therapy. This method not only con-firms reflux using clinical characteristics, but also imple-ments treatment for suspected cases.

Pepsin plays an important role in the development ofmany reflux-related disorders. Pepsin has been posited as areliable diagnostic marker of LPRD7 because it can beeasily detected in airway secretions when gastric reflux hasoccurred for an extended period of time. The recently pub-lished data on the assessment of throat sputum samples hasshowed that pepsin immunoassay is as sensitive and specificfor LPRD as 24-hour double-probe (esophageal and pha-ryngeal) pH monitoring.8 Although pepsin is inactive at pH6.5 and above, it remains stable up to pH 8.0 and can bereactivated when the pH is reduced by a subsequent acidicreflux event. In addition, pepsin is stable for at least 24hours at pH 7.0 at 37°C and retains 79 percent (SD 11%) ofits original activity after re-acidification at pH 3.0.9 Bulmeret al10 reported that laryngeal tissues are essentially resistant

k Surgery Foundation. All rights reserved.

766 Otolaryngology–Head and Neck Surgery, Vol 143, No 6, December 2010

to damage up to pH 4.0 but are damaged when pepsin ispresent in the porcine larynx model. The subglottic mucosaand the vocal folds are damaged more easily than the pos-terior commissure and ventricular mucosa. However, westill have limited information about how pepsin contributesto the laryngeal inflammation expressed by the symptomsand by laryngeal findings in patients.

There were two objectives in this study: 1) to evaluatethe correlation between the severity of clinical characteris-tics and sputum pepsin concentration in two diseases oftenassociated with LPR: reflux laryngitis and OSA; and 2) tocompare the methods and timing of sputum collection.

Materials and Methods

The study protocol was approved by the Committee ofNanfang Hospital. Before initiation of any procedure,signed informed consent was obtained from all participants.

SubjectsWe recruited 56 subjects (27 men, 29 women; mean age37.77 � 10.59 years) enrolled in the Department of Otolar-yngology of Nanfang Hospital from April to October 2007with unspecific laryngopharyngeal symptoms such aschronic cough, throat clearing, globus sensation, andhoarseness. Thirty-six subjects (34 men, 2 women; meanage 40.67 � 7.42 years) registered in the Center for SleepDisorders of Nanfang Hospital during April to June 2008,were likewise recruited as the study’s OSA patients. Allpatients with acute infections of the upper respiratory tract,chronic tonsillitis, chronic sinusitis, malignant tumor, or ahistory of antireflux medical therapy or continuous positiveairway pressure treatment for one month or more wereexcluded. The normal group consisted of 15 healthy sub-jects (9 men, 6 women; mean age 25.07 � 3.24 years).

Study DesignThe laryngitis patients and healthy subjects were asked tocomplete the self-administered nine-item RSI question-naire11 and were examined by videolaryngoscopy to revealany laryngeal mucosal abnormalities. The video recordingswere evaluated using RFS12 by two otolaryngologists, each

Table 1

Characteristics of the groups

Normal(n � 15)

Laryngitis

Reflux laryngitis(n � 32)

Age, yrsMean � SD 25.07 � 3.24 38.75 � 9.96Range 17-30 18-60Sex (M:F) 9:6 16:16

OSA, obstructive sleep apnea.

with more than five years of experience in the field, whoworked independently and were blind to the patient data.The mean score for the subjects was used for analysis of theresults. The laryngitis patients were treated with 20 mg ofomeprazole twice daily for two weeks. A diagnosis of refluxlaryngitis (16 men, 16 women; mean age 38.75 � 9.96years) was confirmed by RSI � 13, RFS � 7, and by theimprovement of characteristic laryngoscopic findings orsymptoms after two weeks of proton pump inhibitor thera-py.6,13 The patients who did not meet the criteria werecategorized into the chronic laryngitis group (11 men, 13women; mean age 36.46 � 11.45 years).

The OSA patients were first asked to complete the Ep-worth Sleepiness Scale (ESS) and the RSI questionnaire.Polysomnography was then performed using a SandmanElite diagnosis recording system (Tyco, Inc., Ottawa, On-tario, Canada). Polysomnography reports established OSAseverity using the apnea-hypopnea index (AHI) and mea-surement of mini SaO2 and mean SaO2 levels. Thirty-sixOSA patients were divided into two groups according to theseverity of their sleep apnea:14 1) the “mild-moderategroup,” AHI 5 to 30 (10 men; mean age 39.30 � 7.60years), and 2) the “severe group,” AHI � 30 (24 men, 2women; mean age 41.19 � 7.43 years) (Table 1).

Throat Sputum CollectionTwo different samples consisting of oral secretions andhypopharyngeal secretions were collected from the laryngi-tis patients and the healthy subjects. The subjects wereasked to spit into a container at least 0.5 mL of sputum to beused for enzyme-linked immunosorbent assay (ELISA).These samples were defined as the oral secretions. Thevideolaryngoscopy and a disposable sputum aspiration cath-eter were inserted into patients through different nostrilsafter superficial anesthesia. Once these catheters had beenadvanced to the sinus piriformis or postcricoid area, thesystem was connected to a vacuum pump to extract thesputum defined as the hypopharyngeal secretions. One hun-dred six oral and hypopharyngeal secretions from 56 laryn-gitis patients and 30 samples from 15 healthy subjects weresuccessfully collected. Unfortunately, two hypopharyngealsecretions from the reflux laryngitis group and four hypo-

(n � 56) OSA group (n � 36)

hronic laryngitis(n � 24)

Mild-moderate(n � 10)

Severe(n � 26)

36.46 � 11.45 39.30 � 7.60 41.19 � 7.4321-59 27-50 26-6311:13 10:0 24:2

group

C

767Wang et al Correlation of pepsin-measured laryngopharyngeal reflux disease . . .

pharyngeal secretions from the chronic laryngitis groupwere lost.

The OSA patients were instructed to produce oral secre-tions only at bedtime and at first waking in the morning,which were defined as bedtime and first-morning secretions,during their polysomnography examinations. Videolaryn-goscopy and hypopharyngeal secretions were not requiredfrom the OSA patients. All of the samples were stored inliquid nitrogen following their collection.

Pepsin AssayThe pepsin assay was performed by a study investigatorwho was blinded to the clinical data. Samples were diluted1:4 (v/v) with 0.1 percent DDT and placed in a 37°C waterbath for 10 minutes to reduce their viscosity, then centri-fuged for 15 minutes at 4°C and 5000 rpm in a centrifuge(Eppendorf AG, Hamburg, Germany). The supernatantswere analyzed by ELISA (E0632h; USCN LIFE, MissouriCity, TX). Those samples with dose-response curves thatwere parallel to the purified pepsin standard curve wereused to calculate pepsin concentrations. Sputum pepsin con-centration was measured in units of ng/mL.

Statistical AnalysisStatistical analyses were performed using SPSS Version15.0 (SPSS, Inc., Chicago, IL). Qualitative variables wereexpressed as a percentage. Continuous variables were ex-pressed as a mean (minimum, maximum), and because theywere not normally distributed, were compared using theKruskal-Wallis H test or the Mann-Whitney test. Correla-tions between variables were assessed using Spearman’srank correlations analysis. Statistical significance was de-termined at P � 0.05. Multiple comparisons used a Bon-ferroni correction to maintain an experiment-wide type Ierror of 0.05 or less.

Results

Pepsin in SputumComparison of pepsin concentrations among reflux

laryngitis, chronic laryngitis, and normal groups. A total

Table 2

Concentration of pepsin in oral secretions and in hypo

Reflux laryngitis

Oral secretion 199.59 (0.00, 3560.53)Hypopharyngeal secretion 116.43 (0.00, 837.68)

Data are given as mean (minimum, maximum). P values are f*Statistically significant.

of 136 samples were tested, including the oral and hypo-

pharyngeal secretions from the above three groups. Pepsinwas found to be present in 93.8 percent (30 of 32), 75percent (18 of 24), and 20 percent (3 of 15) of the oralsecretions in the reflux laryngitis, chronic laryngitis, andnormal groups, respectively. Similar results were observedin the hypopharyngeal secretions, with pepsin present in76.6 percent (23 of 30), 60 percent (12 of 20), and 26.7percent (4 of 15), respectively. The pepsin concentrationsare depicted in Table 2. There were dramatic differencesbetween the three groups regarding pepsin concentration inthe oral secretions (P � 0.000) and in the hypopharyngealsecretions (P � 0.000). Moreover, no significant differencesin pepsin concentrations were found in the oral secretionscompared with the hypopharyngeal secretions (P � 0.136).Therefore, we collected only oral secretions in the OSAstudy.

Comparison of pepsin concentrations between themild-moderate and severe groups of OSA patients. Table 3shows the concentrations of pepsin in a total of 72 oralsecretions in the different OSA groups and at the bedtimeand first-morning collection times. Oral secretions in twodifferent times gave nearly exactly the same positive resultfor pepsin presence: 80.0 percent (8 of 10) positive in themild-moderate group and 80.8 percent (21 of 26) positive inthe severe group. The differences in pepsin concentration inthe first-morning samples between the two groups weresignificant (P � 0.019), but bedtime samples were insignif-icant (P � 0.590). However, the two time points did notshow significant differences (P � 0.845).

ngeal secretions in the different groups (ng/mL)

Chronic laryngitis Normal P value

6.37 (0.00, 47.64) 0.36 (0.00, 3.04) 0.000*10.92 (0.00, 60.81) 0.61 (0.00, 3.75) 0.000*

e Kruskal-Wallis H test.

Table 3

Concentration of pepsin at different collection times

in the OSA groups (ng/mL)

Mild-moderate Severe P value

Bedtime 3.66 (0.00, 16.98) 6.26 (0.00, 55.90) 0.590First morning 1.13 (0.00, 3.63) 5.56 (0.00, 28.10) 0.019*

Data are given as mean (minimum, maximum). P values arefrom the Mann-Whitney test.OSA, obstructive sleep apnea.*Statistically significant.

phary

rom th

768 Otolaryngology–Head and Neck Surgery, Vol 143, No 6, December 2010

Correlation of Pepsin Results, Symptoms, andEndoscopic Laryngeal Findings with LaryngitisReceiver operating characteristic curve analysis was used todiscriminate the diseased cases from the normal cases. Thecriterion value we selected to discriminate between thereflux laryngitis and the normal group was 0.1081. A 93.8percent sensitivity and an 80 percent specificity of pepsin-measured LPRD were compared to our combined diagnosis.A concentration of pepsin greater than 0.1081ng/mL wasdefined as pepsin-positive.

Laryngeal reflux symptoms were classified according tothe RSI (range 0-45). Table 4 shows the scores of symptomsfor both pepsin-positive and pepsin-negative cases. In thepepsin-positive oral secretions, we found a significantlyhigher total score mean value (P � 0.000) of 10.98 (0, 21)points compared with the pepsin-negative (3.95 [0, 15]points). The same result was found for pepsin-positive hy-

Table 4

Scores of reflux symptom index for pepsin-positive an

group

Total scoreHoarsenessClearing throat oftenSensation of something sticking in throatTroublesome or annoying coughBreathing difficulties or choking episodesCoughing after eating or lying downDifficulty in swallowing food, liquids, or pillsSensation of lump in the throatHeartburn, chest pain, indigestion, acid coming up

Data are given as mean (minimum, maximum). P values aresubscores are adjusted by the Bonferroni correction (0.05/9 �*Statistically significant.

Table 5

Reflux finding scores for pepsin-positive and -negative

Peps(n

Total score 8.56Subglottic edema 0.22Ventricular 1.04Erythema/hyperemia 2.20Vocal fold edema 1.42Diffuse laryngeal edema 0.88Posterior commissure hypertrophy 1.44Granuloma/granulation tissue 0.24Thick endolaryngeal mucus 1.20

Data are given as mean (minimum, maximum). P values aresubscores are adjusted by the Bonferroni correction (0.05/8 �*Statistically significant.

popharyngeal secretions, with a mean of 10.44 (0, 21) pointscompared with the pepsin-negative secretions value of 6.04(0, 16) points (data not shown). The scores of clearing throatoften (P � 0.000), troublesome or annoying cough (P �0.002), coughing after eating or lying down (P � 0.003),and sensing a lump in the throat (P � 0.000) in the pepsin-positive group were higher than those in the pepsin-negativegroup.

Endoscopic laryngeal findings were classified accordingto RFS (range 0-26). There was a close correlation betweenthe independent values from the scores of two otolaryngolo-gists, with a correlation coefficient of r � 0.620 (P �0.000). We found a significantly higher total score meanvalue (P � 0.007) of 8.56 (2.5, 18) points for the positive-pepsin oral secretions compared with the value of the pep-sin-negative secretions (5.95 [0.5, 15.5] points) (Table 5).However, no significant differences in the respective RFS

ative cases in the laryngitis group and the normal

Pepsin-positive(n � 51)

Pepsin-negative(n � 20) P value

10.98 (0, 21) 3.95 (0, 15) 0.000*1.20 (0, 5) 0.70 (0, 4) 0.1571.84 (0, 5) 0.40 (0, 2) 0.000*1.57 (0, 5) 0.75 (0, 4) 0.0271.27 (0, 5) 0.25 (0, 2) 0.002*0.08 (0, 2) 0.10 (0, 2) 0.9110.82 (0, 3) 0.10 (0, 2) 0.003*0.00 (0, 0) 0.05 (0, 1) 0.1102.96 (0, 5) 1.30 (0, 4) 0.000*1.22 (0, 4) 0.30 (0, 3) 0.007

the Mann-Whitney test. The statistical significance level of5).

s in the laryngitis group and the normal group

itive)

Pepsin-negative(n � 20) P value

18) 5.95 (0.5, 15.5) 0.007*) 0.05 (0, 1) 0.165) 0.55 (0, 2) 0.033) 1.95 (0, 4) 0.564) 0.95 (0, 3.5) 0.022.5) 0.58 (0, 2.5) 0.069.5) 1.30 (0, 2.5) 0.497) 0.00 (0, 0) 0.083) 0.65 (0, 2) 0.008

the Mann-Whitney test. The statistical significance level of1).

d -neg

from0.005

case

in-pos� 51

(2.5,(0, 2(0, 3(0, 4(0, 4(0, 2(0, 3(0, 2(0, 2

from0.006

769Wang et al Correlation of pepsin-measured laryngopharyngeal reflux disease . . .

subscores was found between the pepsin-positive and -neg-ative groups, and similar results were found in the hypo-pharyngeal secretions (P � 0.006).

Relationships among Pepsin, RSI,and OSA SeveritySpearman’s rank correlation analyses were used to evaluatepotential correlations among pepsin, RSI, and OSA severityin the OSA group. A complete list of the associations testedis presented in Table 6. There was a significant correlationbetween the pepsin levels and OSA severity in first-morningsamples. These cases included the positive correlation be-tween the pepsin level and AHI (P � 0.005), the negativecorrelation between the pepsin level and the level of miniSaO2 (P � 0.003), and the negative correlation between thepepsin level and the mean SaO2 (P � 0.006). For allsubjects, the total score and the subscores of the RSI did notsignificantly correlate with the sleep apnea variables (AHI,mini SaO2, and mean SaO2) or with the pepsin levels (datanot shown).

Discussion

Since Belafsky et al11,12 first reported the use of the RSI andRFS to quantify the symptoms and signs of LPR in 2001, avariety of studies have been conducted to validate the di-

Table 6

Parameters assessed by Spearman’s rank

correlations analysis in the obstructive sleep apnea

groups

Parameters assessedr Correlationcoefficient P value

Concentration of pepsin(bedtime)

vs BMI 0.22 0.201vs AHI 0.34 0.044vs ESS 0.06 0.741vs Minimum oxygen

saturation �0.20 0.240vs Mean oxygen saturation �0.16 0.347

Concentration of pepsin(first-morning)

vs BMI 0.42 0.010vs AHI 0.46 0.005*vs ESS 0.22 0.194vs Minimum oxygen

saturation �0.49 0.003*vs Mean oxygen saturation �0.45 0.006*

P values are from the Spearman’s rank correlation analysis.The statistical significance levels are adjusted by the Bonfer-roni correction (0.05/5 � 0.01).BMI, body mass index; AHI, apnea-hypopnea index; ESS,Epworth Sleepiness Scale.*Statistically significant.

agnostic value of the two scales. However, due to the

nonspecific throat symptoms and signs caused by LPR, theuse of only two scales is not reliable for LPR diagnosis oflow-specificity cases.15 The empirical proton pump inhibi-tor test is useful in the diagnosis of LPR, but the criteria forits dosing, treatment duration, and efficacy parameters havenot been established. Siupsinskiene and Adamonis13 pre-sented a significant improvement in the total symptomsscore, posterior laryngitis score, overall vocal dysfunctiondegree, and self-rated well-being in general for 120 patientstreated with omeprazole for two weeks. They concluded thatshort-term treatment with omeprazole could be used toconfirm the clinically based diagnosis of LPR. Therefore, inthis study, we performed diagnostic test by using 20 mg ofomeprazole twice daily for two weeks. Then, according tothe results of diagnostic test as well as RSI and RFS, LPRcould be diagnosed. In recent years, studies have shown thatthe pepsin in sputum can be used as a fast, effective, andinexpensive method to diagnosis LPR.8,16,17 Sputum or sa-liva is easily obtainable in many patient groups (especiallyin adults, children, and the neurologically impaired).18

ELISA is a sensitive method that has been used for thedetection of pepsin.8 Our study explores the relationshipsamong the concentration of pepsin in sputum and the throatsymptoms and signs caused by LPR.

We have found that the level of pepsin was significantlydifferent among the reflux laryngitis, chronic laryngitis, andnormal groups (P � 0.000). The mean concentration ofpepsin in the reflux laryngitis group was 199.59 ng/mL,which was consistent with the 180 ng/mL reported byKnight et al8 and much lower than the concentration in thegastric juice (269 �g/mL).19 In 75 percent of chronic lar-yngitis patients, however, the mean concentration was lowerthan that of the reflux laryngitis group. Moreover, in termsof the limitation of these diagnoses for LPR, the patientswho belong to the reflux laryngitis group may be confirmedas cases of chronic laryngitis. These chronic cases may leadto the high presence of pepsin. The mean pepsin concen-tration in the normal group was 0.36 ng/mL, which confirmsagain that the pepsin exists under physiological conditions,although at a much lower level. Normal individuals cantolerate many episodes of the reflux of gastric contentswithout suffering esophageal symptoms. Additionally, not-withstanding the variance of pepsin concentration occurringin the mild-moderate and severe groups, the mean pepsinconcentrations were much lower (mild-moderate group:1.13 ng/mL; severe group: 5.58 ng/mL) than the concentra-tion in the reflux laryngitis group. Despite the lower pepsinconcentration in the OSA patients, there was a positivecorrelation between pepsin concentration and AHI in thefirst-morning samples.

Symptoms and signs were compared between the pepsin-positive and -negative groups according to the criterionvalue of oral secretions. We found that the RSI and RFStotal scores in the pepsin-positive group were much higherthan those in the pepsin-negative group. For the subscores

of RSI, clearing throat often, coughing, and sensing a lump

770 Otolaryngology–Head and Neck Surgery, Vol 143, No 6, December 2010

in the throat were more frequent and intense in the pepsin-positive group. These results are similar to those reported byPapakonstantinou et al.20 Nevertheless, there was no dra-matic difference in the subscores of RFS between the pep-sin-positive and -negative groups. Because there were atyp-ical findings in the videolaryngoscopy, the RFS subscorestaken alone were of limited value in the prediction or con-firmation of LPR. Meaningful results could be obtainedonly when we were able to consider these scores together.The absence of a correlation between the RSI and theseverity of OSA by way of the values of the AHI or theSaO2 in our study were also consistent with what Wise etal21 and Valipour et al22 reported. In these studies, there wasno direct correlation between reflux symptoms and OSA.Combined with the positive correlation between pepsin andAHI, we confirmed again that the reflux of stomach contentsexists in OSA. For the lower pepsin concentration, however,there were no typical clinical reflux symptoms in OSA.Therefore, assessment of the relationship between OSA andLPR through symptoms alone is not enough.

Nocturnal reflux is common in patients with OSA.22,23

The physiological antireflux mechanisms (swallowing rate,salivation, the pressure of the upper and lower esophagealsphincters, and gastric emptying) are reduced, and the“heartburn signal” is depressed during sleep.24,25 To com-pare the pepsin concentration in different conditions beforeand after sleep, we collected oral secretions before bedtimeand first-morning samples for the same patient. Using thismethod, we have identified two types of reflux episodes:upright or daytime LPR, and supine or nocturnal LPR. Nosignificant difference was found in the pepsin concentrationof the oral secretions from the two types. However, therewas a significant association among the AHI, mini SaO2,mean SaO2, and pepsin concentration only in the first-morning (but not bedtime) samples. The mechanism for thisassociation is not fully understood. It may be due to thepromotion of reflux and increased arousal by decreasedintrathoracic pressure during apnea episodes coupled withtransient lower esophageal sphincter relaxation triggered bydecreased sleep efficiency. These possibly tend to promotemore pepsin reflux into the larynx during the nighttime.26

Therefore, the pepsin concentration in the morning has amore significant relationship with the severity of OSA.

Conclusions

The concentration of pepsin in sputum appears to be asso-ciated with reflux laryngitis and OSA. A higher level ofpepsin was related to higher total scores of RSI and RFS inlaryngitis. In our opinion, the symptoms of the RSI sub-scores, including clearing throat often, troublesome or an-noying cough, coughing after eating or lying down, andsensing a lump in the throat, are strongly correlated withpepsin-measured LPR. The RFS subscores alone had lim-ited value in the evaluation of LPR. However, no relation-

ship between pepsin levels and RSI was found in OSA.

There are no significant differences of pepsin concentra-tion in the sputum collection methods between the oral andhypopharyngeal secretions or in the sputum collection tim-ing between the bedtime and first-morning secretions, sug-gesting that it is not necessary to catheterize participants toobtain pepsin samples in future studies.

Acknowledgments

The authors thank Sheng-li An, PhD (Department of Biostatistics, SouthernMedical University), for his invaluable contributions in providing statisti-cal guidance, and Dr. Qian Zhong for her help in preparing the manuscript.

Author Information

From the Department of Otolaryngology–Head and Neck Surgery (Drs.Wang, Liu, and Li, and Mr. Liu, Mr. Zeng, and Mr. Wu), Department ofClinical Laboratory (Mr. Yang), and The Center for Sleep Disorders (Mr.Shen), Nanfang Hospital of Southern Medical University, Guangzhou,China.

Corresponding author: Xiang-ping Li, MD, Department of Otolaryngol-ogy–Head and Neck Surgery, Nanfang Hospital of Southern MedicalUniversity, North of Guangzhou Road 1838, Guangzhou, Guangdong510515, China.

E-mail address: lp133402@yahoo.com.cn.

Author Contributions

Lu Wang, drafting the article, acquisition of data, interpretation of data;Xiong Liu, acquisition of data; You-li Liu, acquisition of data; Fang-fangZeng, acquisition of data; Ting Wu, acquisition of data; Chun-li Yang,acquisition of data; Hai-yan Shen, acquisition of data; Xiang-ping Li,design of study, drafting the article, interpretation of data.

Disclosures

Competing interests: None.

Sponsorships: None.

References

1. Koufman JA, Aviv JE, Casiano RR, et al. Laryngopharyngeal reflux:position statement of the Committee on Speech, Voice, and Swallow-ing Disorders of the American Academy of Otolaryngology–Head andNeck Surgery. Otolaryngol Head Neck Surg 2002;127:32–5.

2. Koufman JA. The otolaryngologic manifestations of gastroesophagealreflux disease (GERD): a clinical investigation of 225 patients usingambulatory 24-hour pH monitoring and an experimental investigationof the role of acid and pepsin in the development of laryngeal injury.Laryngoscope 1991;101(4 Pt 2 Suppl 53):1–78.

3. Gurski RR, da Rosa AR, do Valle E, et al. Extraesophageal manifes-tations of gastroesophageal reflux disease. J Bras Pneumol 2006;32:150–60.

4. Abd El-Fattah AM, Abdul Maksoud GA, Ramadan AS, et al. Pepsinassay: a marker for reflux in pediatric glue ear. Otolaryngol Head NeckSurg 2007;136:464–70.

5. Ali Mel-S. laryngopharyngeal reflux: diagnosis and treatment of acontroversial disease. Curr Opin Allergy Clin Immunol 2008;8:28–33.

6. Ford CN. Evaluation and management of laryngopharyngeal reflux.JAMA 2005;294:1534–40.

7. Samuels TL, Johnston N. Pepsin as a marker of extraesophageal reflux.

Ann Otol Rhinol Laryngol 2010;119:203–8.

771Wang et al Correlation of pepsin-measured laryngopharyngeal reflux disease . . .

8. Knight J, Lively MO, Johnston N, et al. Sensitive pepsin immunoassayfor detection of laryngopharyngeal reflux. Laryngoscope 2005;115:1473–8.

9. Johnston N, Dettmar PW, Bishwokarma B, et al. Activity/stability ofhuman pepsin: implications for reflux attributed laryngeal disease.Laryngoscope 2007;117:1036–9.

10. Bulmer DM, Ali MS, Brownlee IA, et al. Laryngeal mucosa: itssusceptibility to damage by acid and pepsin. Laryngoscope 2010;120:777–82.

11. Belafsky PC, Postma GN, Koufman JA. Validity and reliability of thereflux symptom index (RSI). J Voice 2002;16:274–7.

12. Belafsky PC, Postma GN, Koufman JA. Validity and reliability of thereflux finding score (RFS). Laryngoscope 2001;111:1313–7.

13. Siupsinskiene N, Adamonis K. Diagnostic test with omeprazole inpatients with posterior laryngitis. Medicina (Kaunas) 2003;39:47–55.

14. American Academy of Sleep Medicine Task Force. Sleep-related breath-ing disorders in adults: recommendations for syndrome definition andmeasurement techniques in clinical research. Sleep 1999;22:667–89.

15. Park KH, Choi SM, Kwon SU, et al. Diagnosis of laryngopharyngealreflux among globus patients. Otolaryngol Head Neck Surg 2006;134:81–5.

16. Potluri S, Friedenberg F, Parkman HP, et al. Comparison of a sputum/sputum pepsin assay with 24-hour esophageal pH monitoring fordetection of gastric reflux into the proximal esophagus, oropharynx,and lung. Dig Dis Sci 2003;48:1813–7.

17. Krishnan U, Mitchell JD, Messina I, et al. Assay of tracheal pepsin as a

18. Streckfus CF, Bigler LR. Saliva as a diagnostic fluid. Oral Dis 2002;8:69–76.

19. Jones AT, Balan KK, Jenkins SA, et al. Assay of gastricsin andindividual pepsins in human gastric juice. J Clin Pathol 1993;46:254–8.

20. Papakonstantinou L, Leslie P, Gray J, et al. Laryngopharyngeal reflux:a prospective analysis of a 34 item symptom questionnaire. ClinOtolaryngol 2009;34:455–9.

21. Wise SK, Wise JC, DelGaudio JM. Gastroesophageal reflux and la-ryngopharyngeal reflux in patients with sleep-disordered breathing.Otolaryngol Head Neck Surg 2006;135:253–7.

22. Valipour A, Makker HK, Hardy R, et al. Symptomatic gastro-oesoph-ageal reflux in subjects with a breathing sleep disorder. Chest 2002;121:1748–53.

23. Green BT, Broughton WA, O’Connor JB. Marked improvement innocturnal gastroesophageal reflux in a large cohort of patients withobstructive sleep apnea treated with continuous positive airway pres-sure. Arch Intern Med 2003;163:41–5.

24. Pasricha PJ. Effect of sleep on gastroesophageal physiology andairway protective mechanisms. Am J Med 2003;115(Suppl 3A):114S– 8S.

25. Demeter P, Pap A. The relationship between gastroesophageal refluxdisease and obstructive sleep apnea. J Gastroenterol 2004;39:815–20.

26. Ozturk O, Ozturk L, Ozdogan A, et al. Variables affecting the occur-rence of gastroesophageal reflux in obstructive sleep apnea patients.

marker of reflux aspiration. J Pediatr Gastroenterol Nutr 2002;35:303–8. Eur Arch Otorhinolaryngol 2004;261:229–32.