American Journal of Otolaryngology–Head and Neck Medicine and Surgery xxx (2017) xxx–xxx
a r t i c l e i n f o a b s t r a c t
Article history: Objective: To determine if an anti-reflux induction program relieves laryngopharyngeal reflux (LPR) symptoms
Received 3 September 2017 more effectively than medication and behavioral changes alone.
Available online xxxx Study design: Retrospective study.
Setting: Tertiary care academic center.
Keywords:
Subjects and methods: A database was populated with patients treated for LPR. Patients were included in the study
Laryngopharyngeal reflux
group if they completed a two-week anti-reflux program (diet, alkaline water, medications, behavioral modifica-
Cough
Induction diet
tions). Patients were included in the control group if they completed anti-reflux medications and behavioral
Proton pump inhibitors modifications only. Patients completed the voice handicap index (VHI), reflux symptom index (RSI), cough se-
Reflux symptom index verity index (CSI), dyspnea index (DI) and eating assessment tool (EAT-10) surveys and underwent laryngoscopy
for examination and reflux finding score (RFS) quantification.
Results: Of 105 study group patients, 96 (91%) reported subjective improvement in their LPR symptoms after an
average 32-day first follow-up and their RSI and CSI scores improved significantly. No significant differences
were found in VHI, DI, or EAT-10 scores. Fifteen study patients who had previously failed adequate high-dose
medication trials reported improvement and their CSI and EAT-10 scores improved significantly. Ninety-five per-
cent of patients with a chief complaint of cough reported improvement and their CSI scores improved significant-
ly from 12.3 to 8.2. Among 81 controls, only 39 (48%) patients reported improvement after an average 62-day
first follow-up. Their RSI scores did not significantly change.
Conclusion: The anti-reflux program yielded rapid and substantial results for a large cohort of patients with LPR. It
compared favorably with medication and behavioral modification alone. It was effective in improving cough and
treating patients who had previously failed medications alone.
© 2017 Published by Elsevier Inc.
1. Introduction years. Between 1990 and 2001, PPI prescription increased 14 fold, ac-
counting for a significant percentage of healthcare costs [5,6]. A typical
The diagnosis of laryngopharyngeal reflux (LPR) is common in clin- treatment for LPR in clinical practice is a course of twice daily PPI for
ical practice. It was reported to be present in N50% of patients with la- at least 2 months [7]. Many studies have supported the effectiveness
ryngeal complaints at an academic voice center [1]. In a recent survey, of PPIs in treating LPR-related symptoms. One meta-analysis of random-
N60% of a community-dwelling population had either GERD or laryn- ized controlled trials showed that patients treated with a PPI had a sig-
geal symptoms and N20% had both [2]. Symptoms caused by LPR in- nificantly higher response rate and reflux symptom index (RSI)
clude chronic cough, dysphonia, dysphagia, post-nasal drip, globus, improvement than those who received placebo [8]. In a study by Jin et
constant throat clearing, laryngospasm and a multitude of other al., treatment with PPIs improved objective voice measures including
extraesophageal maladies [3,4]. jitter, shimmer, and harmonic-to-noise ratio after 1–2 months treat-
Empiric treatment of LPR with antireflux medication such as proton ment, and maintained results even after 3–4 months [9]. Most studies
pump inhibitors (PPI) has increased in popularity over the past twenty agree that patients must continue their medication regimen for at
least 2–6 months to achieve reduction in symptoms. However, these
studies have been contradicted by others in the literature due to dis-
☆ Presented at Western Medical Research Conference, Carmel, California, USA on Jan. 27,
2017.
crepancies in method of diagnosis, contributing factors, management
⁎ Corresponding author. regimens, and outcome measures that are often subjective and vary
E-mail address: jiyang@llu.edu (J. Yang). widely.
https://doi.org/10.1016/j.amjoto.2017.10.014
0196-0709/© 2017 Published by Elsevier Inc.
Please cite this article as: Yang J, et al, Treating laryngopharyngeal reflux: Evaluation of an anti-reflux program with comparison to medications,
American Journal of Otolaryngology–Head and Neck Medicine and Surgery (2017), https://doi.org/10.1016/j.amjoto.2017.10.014
2 J. Yang et al. / American Journal of Otolaryngology–Head and Neck Medicine and Surgery xxx (2017) xxx–xxx
Though some studies support the use of PPIs for LPR, other studies water (pH N 8) daily [18], and behavioral modifications, including
call into question their efficacy or outline the potential dangers of PPI weight loss, smoking cessation, alcohol avoidance, and eating no less
use over protracted periods of time. First, PPIs have been shown to be than 3 h before lying down [19]. If patients preferred or presented
more beneficial for patients with possible LPR in the setting of typical with PPI 40 mg BID instead of 40 mg qD, their medication regimen
GERD than for those without it [10]. Second, patients may not universal- was maintained. At the end of two weeks, these patients were
ly respond to PPIs: many are PPI resistant or have non-acid reflux [11]. instructed to begin reintroducing foods back into their diet slowly in
Third, the literature does not conclusively support using PPIs to improve order to monitor rebound symptoms and to subsequently determine
LPR symptoms and objective voice measures. A double-blinded, ran- which foods were causing problems so that these could be avoided.
domized controlled trial showed no significant difference between PPI They were asked to return within a month of beginning treatment in
and placebo groups in improving RSI and RFS [7] and, in contrast to Jin order to gauge initial success with the induction diet and to help guide
et al., Hamdan et al. showed that PPI use did not improve acoustic ab- transition to the maintenance phase, if appropriate. We retrospectively
normalities [9,12]. In addition, long term use of PPIs has been associated excluded only patients who failed to follow up within 2 months or who
with such adverse effects as osteoporosis, infections, malabsorption, reported to have not been 100% adherent to the treatment protocol.
malignancy, kidney disease and dementia, sparking great concern in pa- There were no other exclusion criteria for our study group.
tients though proof of causation is largely absent [13–16]. But the pre- Our control group was comprised of the remainder of our patients
scription of anti-reflux medications for LPR treatment continues. who were prescribed a course of high dose PPIs (40 mg qD), or both
The anti-reflux induction diet was introduced by Dr. Koufman in high dose PPIs and H2 blockers (300 mg qHS) with LPR behavioral mod-
2011 [17]. It is comprised of low-acid, low-fat foods to the exclusion of ifications (Table 2) [19]. If patients preferred or presented with PPI 40 mg
all foods and drinks with a pH less than five for a minimum two week BID instead of 40 mg qD, their medication regimen was maintained.
period. The purpose is to provide a basis for what will be a long-term These patients were largely seen prior to the introduction of the anti-
lifestyle change to potentially alter the mechanism and minimize the reflux induction program into our practice and standard follow-up for
effects of LPR. The induction diet ends with transition to a similar but this group was three months, consistent with reports that treatment
less stringent maintenance diet intended to eliminate the need for requires at least 2 months to take effect [7]. We excluded patients who
daily PPIs. This diet/lifestyle approach could provide an alternative for failed to follow up within 3 months in order to bring this group as close
patients refractory to PPIs or who wish to avoid side effects of long as possible to our study group. Patients were also excluded if they report-
term PPI use. In her prospective study of 20 patients who failed PPIs, ed being noncompliant with their medications and behavioral modifica-
19 patients improved on this low acid diet and 3 became completely tions. There were no other exclusion criteria for our control group.
asymptomatic [16]. Our study sought to address whether this induction For both study and control groups, information including routine de-
diet would be effective in reducing LPR symptoms in a larger patient mographics, previous treatment with antireflux medication (dosage
population. We present the outcomes of a regimen combining the in- and duration), medical comorbidities, smoking status, and 24-hour pH
duction diet with anti-reflux medications in a larger group. We compare probe studies was gathered. At each of their clinic visits, patients were
these results with a group who received standard anti-reflux treatment asked to complete the VHI-10 [20] and RSI [21]. Additional question-
in our practice. naires were added at the time the induction program was introduced
and these included the CSI [22], DI [23], and EAT-10 [24]. In addition,
2. Methods an otolaryngologic history and physical exam was completed accompa-
nied by a videostroboscopic exam, which was assigned a Reflux Finding
Loma Linda University IRB granted approval for a retrospective review Score (RFS) [25]. Information about patients' subjective symptom
of patients treated for laryngopharyngeal reflux (LPR) at our academic improvement was collected during their first follow-up visit.
tertiary referral center. A database was assembled by initially including Two subgroups were isolated from the treatment group for further
all patients over age 18 diagnosed with primary LPR from 12/2011 to 6/ scrutiny. The first included all patients in the treatment group who
2016 at the Loma Linda Voice and Swallowing Center (LLVSC). Diagnosis had failed a complete course of 40 mg qD or higher dosage PPI for
was based on the presence of signs and symptoms or pH probes N6 weeks before presenting to our center and completing the LPR in-
(nasopharyngeal pH probes at our center with positive Ryan scores and duction program. Six weeks was chosen in order to exclude patients
adequate symptom correlation or outside reports of positive pH probe who had completed longer courses than the popular 14 day trial. The
results) demonstrating LPR or GERD (with extraesophageal symptoms). second group was comprised of treatment group patients presenting
Patients who had negative pH probe results, and who had other contrib- with the chief complaint of cough who had also completed the CSI ques-
uting pathology (radiation, vocal fold lesions, sinonasal pathology, airway tionnaire pre- and post-induction program. Pre-and post-induction pro-
stenosis, etc.) were excluded. Patients who were simultaneously treated gram questionnaire scores and symptom results were analyzed
for or found to have allergic rhinitis, glottic insufficiency, or vocal fold separately for the above two cohorts.
atrophy were also excluded as these other treatments could cause the The RFS was assigned to every exam collected from patients in the
LPR-directed treatments to appear more effective. We divided all patients treatment group. They were not collected from the control group be-
meeting criteria into two groups: those who were prescribed the LPR cause most exams were not available for review. Two different attend-
induction program and those who were prescribed only anti-reflux ing laryngologists performed the scoring blinded to the patient, date
medications and behavioral modifications. of exam, and treatment period. Using only patients who had both pre-
Patients prescribed the LPR induction program (Table 1) were and post-induction program exams within two months of each other,
included in our study group. The LPR induction program consists of a overall change in RFS, inter-rater reliability, and intra-rater reliability
two-week induction diet [17], high dose anti-reflux medications (PPI (20% blinded repeat grading) were calculated.
40 mg qD and/or H2 blocker 300 mg qHS), with at least 16 oz of alkaline
Table 2
Behavior modifications [19].
Table 1
Anti-reflux program. Weight loss
Smoking cessation
Two-week induction diet [17] Alcohol avoidance
High dose anti-reflux medications (PPI 40 mg qD and/or H2 blocker 300 mg qHS) Minimizing tight clothing/belts
At least 16 oz of alkaline water (pH N 8) daily [18] Eating no less than 3 h before lying down
Behavior modifications (Table 2) Taking PPI 30–60 min before meals
Please cite this article as: Yang J, et al, Treating laryngopharyngeal reflux: Evaluation of an anti-reflux program with comparison to medications,
American Journal of Otolaryngology–Head and Neck Medicine and Surgery (2017), https://doi.org/10.1016/j.amjoto.2017.10.014
J. Yang et al. / American Journal of Otolaryngology–Head and Neck Medicine and Surgery xxx (2017) xxx–xxx 3
gorical variables were calculated. Univariable comparisons between Pre-treatment Post-treatment P-value
pre-and post-treatment for the treatment group, control group, high Study group (N = 105)
dose PPI trial group, cough group, and pH probe group were performed VHI 7.89 6.7 0.077
with paired t-tests for continuous variables. A significance level of 0.05 RSI 19.74 14.38 2.07E−10*
was used for two group comparisons. Tukey adjusted post hoc test CSI 9.58 7.47 0.008*
DI 9.4 8.02 0.31
was used to calculate the percent agreement between patients' reported
EAT-10 9.24 7.81 0.08
improvement and their RSI score improvement.
Failed prior PPI (N = 16)
VHI 10.81 9.88 0.49
RSI 20.13 16.5 0.15
3. Results CSI 14.73 10.22 0.04*
DI 15.29 14.67 0.54
Two hundred and sixty nine patients presented to the LLVSC with EAT-10 8.45 7.17 0.02*
symptomatic LPR and were given the induction program between 10/ Cougher (N = 37)
2014 and 6/2016. Of those, the following patients were excluded from VHI 9.34 8.26 0.27
the study group: 130 patients who had no follow up within two months, RSI 20.7 16.42 0.001*
22 patients who followed up within two months but failed to complete CSI 12.29 8.16 0.005*
DI 11.52 9.52 0.11
the induction program, 3 patients who had negative pH probe results,
EAT-10 9.39 8.7 0.16
and 9 patients who had other pathological contributors for their symp-
toms (Table 3). Among the remainder of the study group who were in- Control group (N = 81)
VHI 9.93 12.31 0.006*
cluded, 15 underwent pH probe testing, the majority at our Center, with
RSI 17.43 17.35 0.989
positive Ryan scores and symptom correlation. Twenty-two patients
Abbreviations: VHI, voice handicap index; RSI, reflux symptom index; CSI, cough severity
who followed up within two months but reported b 100% compliance
index; DI, dyspnea index; EAT-10, eating assessment tool.
with the program had an average age of 61 (range = 32–84, median Note: paired t-test is used and a P-value of 0.05 is selected for significant difference.
= 62) and an average BMI of 28 (range = 18–49, median = 27). Four
of them were male and eighteen were female. Among them, eleven
had respiratory comorbidities (COPD, bronchitis, asthma, allergies), Thirty-seven patients in the treatment group presented with a
one was an active smoker, five were diagnosed with GERD, two had di- chief complaint of cough. Twenty reported cough symptoms lasting
abetes, and one had anxiety. The RSI score for this cohort decreased in- for N1 year, 16 for N 8 weeks and less than a year, and 1 for
significantly from 18.3 to 15.8 (P = 0.08). 5 weeks. Twenty-three of 37 had tried some form of therapy: 16
Of the 105 patients who completed the program and were included had PPI ± H2 blockers, 2 had gabapentin or amitriptyline, 2 had
in the study group, 28 were male and 77 were female. Their mean age over-the-counter cough suppressants, and 3 had tried herbal reme-
was 60 (range = 17–84, median = 63) and mean BMI was 29 (range dies. Thirty-five patients reported subjective improvement with the in-
= 17–45, median = 28). The average first follow up after receiving duction program after a mean 31 days follow-up (range = 15–63,
the induction program was 32 days (range = 7–63, median = 28). In median = 26). Their CSI scores improved significantly from 12.29 to
just this period of time, 96 (91%) patients reported significant subjective 8.16 (Table 4).
improvement or complete resolution of their LPR symptoms. Nine (9%) Two hundred and four patients presented at LLVSC with symptom-
patients did not endorse improvement, one of whom had success upon atic LPR and were given a course of high dose PPIs (40 mg qD or BID),
a second trial of the same program. There were significant differences in or both high dose PPIs and H2 blockers (300 mg qHS) with LPR behav-
RSI and CSI scores pre-and post-induction program for patients who ioral modifications (Table 2) between 12/2011 and 6/2016. Of those, the
completed the induction program (Table 4, Fig. 1). No significant differ- following patients were excluded from the control group: 99 patients
ences were found in their VHI, DI, or EAT-10 scores. The RFS scores did who had other pathological contributors for their symptoms or were
not significantly change after treatment over this time period. Inter- non-compliant with medications, 21 patients who had no follow up
rater reliability was fair and intra-rater reliability was both significant within three months, and 3 patients who had negative pH probe results.
and correlated (intraclass correlation 0.67; 95% CI 0.3–0.87). Our control group was comprised of 81 patients (40% of our initial
Of note, 79 (75%) patients in the treatment group had been on some cohort), 19 male and 62 female. They had a mean age of 59 (range =
form of anti-reflux medications (H2 blocker, PPI or both) prior to begin- 17–88, median = 61) and mean BMI of 29 (range = 18–44, median
ning this regimen. Specific regimen information was available for six- = 29). After an average of 62 days first follow up (range = 14–93, me-
teen of these patients who reported they had completed at least dian = 63), 39 (48%) patients reported subjective improvement of LPR
6 weeks of high-dose PPI ± H2 blocker but failed to improve symptom- symptoms. However, RSI scores did not change significantly in this time
atically prior to beginning the anti-reflux induction program. Of those, period and VHI significantly worsened from a mean of 9.93 to 12.31
15 (94%) reported subjective improvement and one reported no im- (Table 4, Fig. 1). Of the 42 who reported no subjective improvement
provement after an average of 38 days follow-up (range = 15–44, me- with medications, they had an average of 57 days follow-up (range =
dian = 37). This group showed significant improvement in CSI and 14–93, median 63). Their VHI score worsened significantly (P = 0.02)
EAT10 scores and none in VHI, RSI, or DI (Table 4). from 13.7 to 16.0 and their RSI score worsened insignificantly (P =
0.14) from 20.6 to 21.4. As our clinic no longer had access to recorded
videostroboscopic exams from the control group, we were unable to de-
Table 3
Additional diagnoses for patients excluded from the study group. termine a cause for this score increase. There were not enough CSI, DI, or
EAT10 data gathered for statistical analysis.
Allergies, rhinitis, post-nasal drip
Tukey adjusted post hoc test showed no significant demographic dif-
Esophageal pathology: eosinophilic esophagitis, cricopharyngeal web, esophageal
stricture ferences (age, gender disparity, BMI) between study and control groups.
Neurogenic cough In addition, the percent agreement between patients' subjective im-
Muscle tension dysphonia provement and their improvement in RSI scores was significantly corre-
Hyoid bone syndrome lated in both the study and control groups, demonstrating consistency
Vocal fold paralysis
between the two measures (Table 5).
Please cite this article as: Yang J, et al, Treating laryngopharyngeal reflux: Evaluation of an anti-reflux program with comparison to medications,
American Journal of Otolaryngology–Head and Neck Medicine and Surgery (2017), https://doi.org/10.1016/j.amjoto.2017.10.014
4 J. Yang et al. / American Journal of Otolaryngology–Head and Neck Medicine and Surgery xxx (2017) xxx–xxx
Fig. 1. Survey scores for study and control groups pre-treatment and post-treatment.
Please cite this article as: Yang J, et al, Treating laryngopharyngeal reflux: Evaluation of an anti-reflux program with comparison to medications,
American Journal of Otolaryngology–Head and Neck Medicine and Surgery (2017), https://doi.org/10.1016/j.amjoto.2017.10.014
J. Yang et al. / American Journal of Otolaryngology–Head and Neck Medicine and Surgery xxx (2017) xxx–xxx 5
Please cite this article as: Yang J, et al, Treating laryngopharyngeal reflux: Evaluation of an anti-reflux program with comparison to medications,
American Journal of Otolaryngology–Head and Neck Medicine and Surgery (2017), https://doi.org/10.1016/j.amjoto.2017.10.014
6 J. Yang et al. / American Journal of Otolaryngology–Head and Neck Medicine and Surgery xxx (2017) xxx–xxx
[32] Johnston N, et al. Pepsin and carbonic anhydrase isoenzyme III as diagnostic markers [34] Johnston N, et al. Effect of pepsin on laryngeal stress protein (Sep70, Sep53, and
for laryngopharyngeal reflux disease. Laryngoscope 2004;114(12):2129–34. Hsp70) response: role in laryngopharyngeal reflux disease. Ann Otol Rhinol
[33] Gill GA, et al. Laryngeal epithelial defenses against laryngopharyngeal reflux: inves- Laryngol 2006;115(1):47–58.
tigations of E-cadherin, carbonic anhydrase isoenzyme III, and pepsin. Ann Otol [35] Johnston N, et al. Activity/stability of human pepsin: implications for reflux attribut-
Rhinol Laryngol 2005;114(12):913–21. ed laryngeal disease. Laryngoscope 2007;117(6):1036–9.
Please cite this article as: Yang J, et al, Treating laryngopharyngeal reflux: Evaluation of an anti-reflux program with comparison to medications,
American Journal of Otolaryngology–Head and Neck Medicine and Surgery (2017), https://doi.org/10.1016/j.amjoto.2017.10.014