Subjective, laryngoscopic, and acoustic measurements of laryngeal reflux before and after treatment with omeprazole

Journal of Voice

Vol. 10, No. 4, pp. 410-.418 © 1996 Lippincott-Raven Publishers, Philadelphia

Subjective, Laryngoscopic, and Acoustic Measurements of Laryngeal Reflux Before and After Treatment with Omeprazole Gary Y. Shaw, Jeff P. Searl, Jeff L. Young, and Philip B. Miner University of Kansas Medical Center, Kansas City, Kansas, U.S.A.

Summary: Laryngeal manifestation of gastroesophageal reflux is felt to be prevalent in our society. In general, diagnosis has been based primarily on symptoms. Historically, additional testing included laryngoscopy, barium swallow, manometry, and more recently, single- and double-probe pH monitoring. We evaluated 68 patients who were symptomatically suggestive of having reflux laryngitis. We administered surveys grading their symptoms. All patients underwent standardized videolaryngostroboscopic evaluation and computerized acoustic analysis. Patients then underwent a uniform therapy of dietary restrictions and omeprazole, a hydrogen ion inhibitor, for 12 weeks. Patients were then retested. This regimen demonstrated an 85% success of relieving symptoms. Utilizing the new laryngoscopic grading system, improvement was found to be statistically significant in improvement of all findings except granulomas. In patients with the pretherapy complaint of hoarseness, acoustic measures of jitter, shimmer, habitual frequency, and frequency range all showed significant improvement. The authors conclude that in patients with symptomatic reflux laryngitis, standardized videolaryngoscopy and, if hoarse, acoustic analysis are useful exam techniques to aide diagnosis and monitor therapy. Anti-reflux therapy with omeprazole is effective and improvement can be objectively demonstrated with the techniques described. Key Words: Reflux laryngitis--Diagnosis--Treatment.

Gastroesophageal reflux disease (GERD) was originally identified by Cherry and Margulies in 1968 as an etiologic factor in laryngeal disease (1). Yet long before this landmark report, unrecognized laryngeal manifestations of reflux disease were detected by Virchow in 1880, who coined the term "pachydermia verrucosa laryngis" to describe the annular epithelial overgrowth centered on the posterior glottis. Although he felt it to be related to gross vocal abuse, this later would be recognized as pathognomonic for laryngeal reflux. Diagnostic

studies had traditionally been either radiologic based, such as barium esophagram, acid barium, and scintography, or physiologic based, such as the acid perfusion test, manometry, and esophageal pH monitoring. Despite the significant cost and inconvenience of these tests, none of them were definitive in helping to diagnose laryngeal manifestations of GERD. Recently dual-probe pH monitors have assisted in diagnosing patients with laryngeal reflux (2,3). However, reports on its sensitivity vary (4). Much recent attention has been focused on the laryngeal symptoms of GERD. It is surmised that hoarseness, for example, is the most common head and neck symptom (5). Despite many reports noting the presence of reflux-related laryngeal disease and its many symptoms, few studies have attempted s y s t e m a t i c a l l y and o b j e c t i v e l y to d e s c r i b e

Accepted November 17, 1995. Address correspondence and reprint requests to Dr. G. Y. Shaw, Department of Otolaryngology, University of Kansas Medical Center, Kansas City, KS 66160, U.S.A.

410

L A R YNGEAL R E F L U X BEFORE A N D A F T E R O M E P R A Z O L E

these signs and symptoms. Utilizing objective measures should facilitate evaluation of treatment regimens. The goals of this study were (a) to identify the common symptoms associated with GERD, (b) to describe objectively the laryngoscopic findings utilizing a standard scale, (c) to identify whether acoustic tests are affected by GERD, and (d) to measure the effect of 3 months of reflux treatment utilizing our symptom survey, laryngoscopic grading, and acoustic analysis. MATERIALS AND METHODS Sixty-eight patients were identified as having laryngeal symptoms suggestive of GERD. There were 22 males and 46 females, whose ages ranged from 21 to 83 years, with a mean of 47 years. They were each administered a symptom questionnaire that inquired whether they had experienced the following: heartburn, foul taste, throat clearing, sensation of something in throat, cough, throat pain, swallowing problems, and hoarseness. They further detailed the frequency of the symptom. Frequency of symptoms was assigned a rating on the following point scale: never (0), rarely (1), occasionally (2), and frequently (3). The subjects then underwent videolaryngoscopy utilizing a 70 ° laryngoscope (Kay Model 9105) and halogen light source (Bruel and Kjaer Model 4914). Images were captured using a CCD camera (Toshiba) coupled to a super VHS videorecorder (Mitsubishi Model BV-1000) and viewed on a high-definition monitor (Sony PVM 1343MD). A permanent image was obtained via a digital recorder (Sony Mavigraph 5000). Evaluation of the laryngoscopic image was performed independently by the authors. A standardized scoring system was utilized, with particular emphasis on the posterior glottis. Graded features included erythema, edema, nodularity, ulceration, and granuloma. Each category was rated absent (0), mild (1), moderate (2), or severe (3). Scores from the two observers were then averaged. Enrolled patients then underwent computerized acoustic analysis. Audio tape recordings were done in a s o u n d p r o o f room. A headset microphone (Telex Model PH-91) was positioned at a constant mouth-to-microphone distance of 2 in. The microphone was coupled to an audio recorder (Marantz Model PMD 201). The audio signal was digitized into a computer (Zenith 386/25) using an AD/DA converter (Data Translation DT2719) and acoustic

411

analysis software (C-Speech, P. Menclivic, Version 3.0). Parameters measured included perturbation of frequency (jitter) and amplitude (shimmer), signalto-noise ratio, modal fundamental frequency, and pitch range. Patients were then administered a therapeutic regimen that entailed the following. (a) Avoid eating or drinking 2 h prior to lying down. (b) Avoid all tobacco products. (c) Avoid fried foods, chocolate, citrus, carbonated beverages, alcohol, mints, onions, garlic, chewing gum, hard candies, coffee, and all caffeinated products. (d) Incline the head of the bed 6 to 8 in. above the feet. If this is not possible (i.e., water bed), obtain a foam bed wedge. (e) Avoid tight clothing. The following medications were prescribed for all patients: (I) Prilosec (omeprazole), 20 mg b.i.d., and (2) Gaviscon, 30 ml p.o. 1 h after meals and 2 h prior to bed. If the patient has a history of bloating, then Propulcid (cisapride), 10 mg p.o.q.i.d., was also prescribed. After 12 weeks of treatment patients again underwent the same evaluation as pretherapy (i.e., symptom survey, videolaryngoscope, and acoustic analysis). Pretherapy evaluations were then compared to posttherapy evaluation. Statistical analysis was employed to determine significant differences (see below). Statistical management Preliminary datum analysis indicated that videolaryngoscopic data were not normally distributed. This, along with the qualitative nature of the ratings, prompted the use of nonparametric statistical analysis. The Sign test was used to evaluate these data. The Sign test is applicable when evaluating paired data and the direction of the change in the paired data is of interest. Essentially, the test evaluates the number of positive changes and negative changes and, given a particular sample size, gives the probability of having that number of positive (or negative) changes. When N is >35, the normal approximation to the binomial distribution can be used. Standard scores (z scores) are appropriate when employing this distribution. The z score indicates the relative position of any value in terms of the number of standard deviations that the value is away from the mean. The probability of obtaining specific z scores under the standard normal distribution can be determined from standard normal distribution tables. The same procedure was used for Journal of Voice, Vol. 10, No. 4, 1996

412

G. Y. S H A W E T A L .

evaluation of the patient symptom questionnaire data. Acoustic analysis data did meet assumptions regarding normal distribution and homoscedasticity. A series of paired t tests was used to evaluate these data. Comparisons were made of all patients' pretherapy versus posttherapy results. Similar comparisons were made in just those patients who complained of either occasional or frequent pretherapy hoarseness. RESULTS

10o

PRE POST

|!

90 82 8O 90 70 Percent of

60

Subjects

50

I 65

48

40 30

20 l0 0 Heart-

Bad Throat Globus Taste Clearing

burn

Cough

Dys ph2gia

s~'ere 3

Odyna- Hoarsephagia n ~ s

"1¢ PRE 2,48

•

2

POST

I 66

I 84

I.

Severity I

1.17 /

1.22

0.69

0.2S

1.19

0 none

0

, Heartburn

j Bad

,

,

Throat Cl~ring

Clabtts

, Cough

, Dysphsgia

, Odynophagia

, HoarseheSS

FIG. I. A. Bar graph representing results of patient symptom survey (frequency of occurrence). Light bars are pretherapy; dark bars, posttherapy. B. Graph representing group mean ratings of symptom severity pretherapy (stars) and posttherapy (circles).

Journal of Voice, Vol. 10, No. 4, 1996

Heartburn Bad taste Throat clearing Globus sensation Coughing Dysphagia Odynophagia Hoarseness

z score

p

Mean improvement

6.421 5.317 6.940 6.261 4.199 5.730 3.742 5.218

0.0000 0.0000 0.0000 0.0000 0.0000 0.0000 0.0020 0.0000

0.39 0.60 0.95 0.84 0.44 0.44 0.16 0.65

Results of Sign test demonstrate the probability that symptom improvement with therapy was significant.

Symptom surveys (see Figs. 1A and B and Table 1) All 68 patients completed symptom surveys. Pretherapy the most common complaint was frequent throat clearing (90%). This was followed by globus sensation (82%), hoarseness (65%), dry cough (53%), dyspepsia (48%), dysphagia (37%), heartburn (34%), and finally, odynophagia (4%). Posttherapy mean symptom scores for each individual symptom were compared to pretherapy mean scores. In every category except odynophagia, a statistically significant improvement was appreciated. Group mean improvement, as measured A

T A B L E 1. Sign test results and group mean improvement f o r each patient symptom category

in scale ratings, was greatest for the measures of throat clearing (0.95-scale point improvement), globus sensation (0.84 scale point), hoarseness (0.65 scale point), bad taste (0.60 scale point), dysphagia (0.44 scale point), and heartburn (0.39 scale point). Videolaryngoscopy (see Figs. 2A-N and 3A and B and Table 2) All 68 patients underwent videolaryngoscopy. The previously described grading system was employed. Mean scores of each individual category were calculated. Pretherapy scores were then compared to posttherapy scores. Statistically significant improvement was seen in all categories except granulomas. Examination of the group mean rating revealed that posterior glottic edema, nodularity, and erythema were the most severe and frequently occurring findings. These same subjects had the most significant improvement posttherapy. Ulceration was a relatively infrequent finding pretherapy (50%) but, again, posttherapy showed a significant improvement. Only three patients had granulomas, and 12 weeks of therapy showed improvement in only one of these patients. Acoustic analysis (see Figs. 4a and b and Table 3) All 68 patients underwent computerized acoustic analysis recording perturbation in amplitude (shimmer), frequency (jitter), signal-to-noise ratio, modal fundamental frequency, and pitch range. Pretherapy results were compared to posttherapy results. Despite the improvement, trends noted that no individual acoustic result differed significantly from pretherapy, when all patients' group mean scores were considered. However, when patients who complained of hoarseness pretherapy were then specifically selected and their pre- and posttherapy results were evaluated, a statistically significant im-

LAR YNGEAL REFLUX BEPORE AND AFTER OMEPRAZOLE

413

A,B

C,D

I i

•~ , ' . ~ .

•

,.

'7

, :,

.k

I

FIG. 2. A. Illustration demonstrating mild posterior glottic nodularity. B. Laryngographic image demonstrating mild posterior glottic nodularity. C. Illustration demonstrating severe posterior glottic nodularity. D. Laryngographic image demonstrating severe posterior glottic nodularity. E. Illustration demonstrating moderate posterior glottic erythema. F. Laryngographic image demonstrating moderate posterior glottic erythema. G. Illustration demonstrating mild posterior glottic edema. H. Laryngographic image demonstrating moderate posterior glottic edema. I. Illustration demonstrating severe posterior glottic edema. J. Laryngographic image demonstrating severe posterior glottic edema. K. Illustration demonstrating moderate posterior glottic ulceration. L. Laryngographic image demonstrating moderate posterior glottic ulceration. M. Illustration demonstrating severe posterior glottic granuloma. N. Laryngographic image demonstrating moderate posterior glottic granuloma.

provement was appreciated. In this group of subjects a statistically significant improvement (p < .05) was seen in jitter, shimmer, and frequency range. Additionally, it was noted in this group of subjects that modal fundamental frequency increased from a mean of 162 Hz pretherapy to 212 Hz posttherapy.

DISCUSSION

The causal relationship between certain laryngopharyngeal lesions and GERD was first proposed by Cherry et al. in 1968 (1). They clinically observed chronic symptoms, such as hoarseness and globus pharyngeus, resolve with prolonged antireJournal of Voice, Vol. IO, No. 4, 1996

E,F

h

G,H

I f

I,.I

I

d FIG. 2. Continued.

LAR YNGEAL R E F L U X BEFORE A N D AFTER O M E P R A Z O L E

415

K,L

M,N

i

FIG. 2. Continued.

flux therapy. Further, their animal studies, in which gastric juices were applied to the posterior glottis, appeared to reproduce the posterior glottic lesions seen in their patients (6). Since that time, a relative controversy has evolved on whether GERD truly is an etiologic factor in many laryngopharyngeal symptoms and physical signs (4,7). The controversy stems from two basic problems: diagnosis and etiology. To date, despite a plethora of diagnostic tests, including radiologic, scintographic, acid provoca-

tion, manometric, single- and dual-probe pH monitoring, and others, there does not appear to be a test that repeatedly detects direct reflux in the area of the laryngopharynx in patients for whom there is a high clinical suspicion. As an example, dual-probe pH monitoring, which is felt by some to be the current gold standard, has had a variably reported sensitivity of 78.8% (2) to 17.5% (4). Additionally, even proponents of dual-probe monitoring have noted false readings from the upper pharyngeal probes (8). Currently there are two schools of thought on the Journal of Voice, Vol. 10, No. 4, 1996

416

G. Y. S H A W E T A L .

A

PRE POST

I00

80

78

70 Percent of Subjects Exhibiting Finding

70

6O

50 40

37

2O

Nodularity

B

Erythema

Ulceration

Edema

Granuloma

severe Nodularity Erythema . . . .

2.2 2.1

"'.. ~ . . --. ~. "

1.9

Ulceration . . . . . . . . Edema ................. Granuloma

~

~

Group Mean

Rating

1.2 1.1 ~...

"~" 1.0, 1.0

0.6 0.2

0.3 none

0 PRE

POST

FIG. 3. A. Bar graph demonstrating videolaryngoscopic findings pretherapy (light bars) and posttherapy (dark bars). B. Graph demonstrating the decrease in group mean severity of all laryngoscopic findings with treatment.

etiologic basis of laryngoesophageal manifestations of GERD. The first is the concept that direct exposure of the posterior glottis to gastric contents causes lesion to develop, which produce the typical symptoms. The second theory proposes that vagaily mediated reflex occurs when the lower esophagus is exposed to gastric acid. The reflex stimulates abnormal muscle contraction in the upper TABLE 2. Sign test results for videolaryngoscopic findings

Nodularity Erythema Ulceration Edema Granuloma

z score

p

6.634 6.857 5.635 6.936 1.002

0.000 0.000 0.002 0.000 0.201

Sign test demonstrates the high statistical probability of laryngoscopic improvement for all parameters measured except granuloma. Journal of Voice, Vol. 10, No. 4, 1996

aerodigestive tract, along with excess salivation and bronchospasms (9). A third possibility is a combination of the two. Despite the diagnostic and etiologic controversy, clinical studies and experience support the fact that many patients with typical laryngopharyngeal symptoms and signs of supposed GERD improve with antireflux therapy (10,11). While many of these studies described symptom improvement, few attempted to analyze acoustic parameters objectively. Fewer still examined laryngoscopic findings pre- and posttherapy. No previous study, of which the authors are aware, has utilized standard grading for laryngoscopic findings. The choice of omeprazole as the predominant antireflux medication and the length of time (12 weeks) was based on the work of Kamel et al. (12) and Koufman (13). They stressed that gastric pH is raised to - 4 . 0 with H 2 antagonists. At that pH some gastric enzymes (e.g., pepsin) still have activity. Omeprazole specifically inhibits the final enzymatic step in the formation of hydrochloric acid in the parietal cell. Its suppression of gastric acid production is near-complete for - 1 6 h. It is therefore dosed twice a day. It should be stressed that, at the time of writing, this application is considered an-off label use of the drug. Nevertheless, its efficacy in treating this disorder is evident. However, if the disease has been long-standing and the tissue damage is severe, then it is to be expected that a certain percentage may not improve. This study demonstrated that physical improvement in posterior glottic tissue lags well behind acoustic and symptomatic improvement. The poor response of contact granuIomas to intensive therapy may point to an alternative etiology to this finding. It should be stated emphatically that other reflux protocols could be employed with success. H 2 antagonists and antacids are usually successful, with improvement in - 7 0 % of patients with laryngeal reflux (13). As stated above, the choice of omeprazole was to provide what is currently felt to be the most effective antireflux treatment, especially for recalcitrant cases of reflux laryngitis. The purpose of this study was to determine if patients with laryngopharyngeal s y m p t o m s of GERD can be objectively described by acoustic and laryngoscopic means pre- and posttherapy to determine improvement. Ideally, this would have been performed in a double-blind fashion with placebo. In fact this study is currently being designed at our institution in a cooperative effort between the De-

LAR YNGEAL REFLUX BEF,ORE AND AFTER OMEPRAZOLE A,B

.10

JITTER

SHIMMER

IO

.09

20

Hoarse Subject Group

SIGNAL/NOISE

Non-Hoane Subject Grou

300

300

250

250

19

.08

\ I ,, ,,\.07o31

.07

8 ,

7.19 ~

7.82

~ " 6.82

.06

\.06641

.05

4

.03

3

18

/

17

16

6 5

.04

15 L5.18 / / ./ 14 13.91 13

17.33

F0 (Hz)

15.40

200

PRE ....

POST

PRE

POST

PRE

I

L

162 150

POST

HOARSE NON-HOARSE

/

/

J

3.12

200

/

150

100

I00

PRE Hoarse Subject Group

POST

i!2 ~ ~ ~ -- " 207

L PRE

Non-Hoarse Subject Group

5OO

POST

500 (295)

(289)

400

400

(249)

(215) F0 (ritz)

417

300

300

200

200

I00

FIG. 4. A. G r a p h s d e m o n s t r a t e the effect of t h e r a p y on the group mean acoustic m e a s u r e m e n t s of jitter, s h i m m e r , and signal-to-noise ratio. While g r o u p m e a n s c o r e s o f all patients showed improvement, only patients with the pretherapy complaint of h o a r s e n e s s (solid line) d e m o n s t r a t e d a statistically significant (p < 0.05) i m p r o v e m e n t in jitter and s h i m m e r . B. G r a p h d e m o n s t r a t e s the increase in group m e a n habitual f u n d a m e n t a l frequency obtained with therapy. Again, note that statistical imp r o v e m e n t was seen only for subjects with h o a r s e n e s s as a pretherapy complaint. C. G r a p h d e m o n s t r a t e s the increase in group m e a n f u n d a m e n t a l f r e q u e n c y range pre- and p o s t t h e r a p y in hoarse and n o n h o a r s e subjects.

10O PRE

POST

PRE

POST

partment of Otolaryngology and the Department of Gastroenterology. N e v e r t h e l e s s , our findings clearly demonstrate a statistically significant improvement in symptoms and laryngoscopic and acoustic parameters using the techniques described. CONCLUSION For nearly 30 years laryngopharyngeal manifestations of GERD have been described. The dif-

ficulty has been causally relating GERD to these physical signs and symptoms. Until a reliable diagnostic test exists, clinicians must maintain a high degree of suspicion that described symptoms and iaryngoscopic findings are attributable at least in part to GERD. Medical treatment should be promptly promptly instituted, with the understanding that it may take a while before resolution occurs. Resolution of symptoms, particularly hoarseness and laryngoscopic findings such as edema, erythema, and nodularity, have been shown to be

TABLE3. Paired t test results f o r acoustic measures Hoarse group (df = 43)

N o n h o a r s e group (df = 23)

t value

Significance

t value

Significance

3.41 3.09 1.74 4.05 3.78 4.41 1.94

0.0081 0.0096 0.078 i 0.0021 0.0043 0.0009 0.986

1.26 0.86 I. 14 1.29 1.74 1.03 1.35

0.1301 0.2008 0.1772 0.1302 0.0562 0.1853 0.1568

F0 High Modal Low Range Jitter Shimmer Signal noise

Paired t test s h o w s the probability of a statistically significant i m p r o v e m e n t in acoustic m e a s u r e s . Only in patients with a pretherapy complaint of h o a r s e n e s s was significant i m p r o v e m e n t seen, with d e c r e a s e s in jitter and s h i m m e r and increases in pitch range and habitual f u n d a m e n t a l frequencies.

Journal of Voice, Vol. lO, No. 4, 1996

418

G. Y. S H A W ET AL.

objectively documentable utilizing acoustic testing and videolaryngoscopy, respectively. It is hoped that others will incorporate these techniques to determine the efficacy of therapy used to treat this disorder. REFERENCES I. Cherry J, Margulies S. Contact ulcer of the larynx. La~.ngoscope 1968;78:1937--40. 2. Wiener G J, Koufman JA, Wu W, et al. Chronic hoarseness secondary to gastroesophageal reflux disease: documentation with 24-H ambulatory pH monitoring. Am J Gastroenterol 1989;84:1503--8. 3. Katz PO. Ambulatory esophageal and hypopharyngeal pH monitoring in patients with hoarseness. Am J Gastroenterol 1990;85:38--40. 4. Wilson JA, White A, Maran AGD, et al. Gastroesophageal reflux and posterior laryngitis. Ann Otol Laryngol 1989;98: 405-10. 5. McNally PR, Maydonovitch CL, Prosek RA, et al. Evalua-

Journal of Voice, Vol. 10, No. 4, 1996

6. 7. 8. 9. 10. I I. 12.

13.

tion of gastroesophageal reflux as a cause of idiopathic hoarseness. Dig Dis Sci 1989;34:1900.-4. Delahunty JE, Cherry J. Experimentally produced vocal cord granuloma. Laryngoscope 1969;78:1941-7. Fraser AG. Review article: Gastroesophageal reflux and laryngeal symptoms, Ali Pharm Ther 1994;8:265-72. McNally PR, Maydonovitch CL, Prosek RA, et al. Hoarseness and gastroesophageal reflux: What is the relationship? Gastroenterology 1990;98:1717-9. Meirowitz RF, Sudhir KD. Esophageal acid perfusion induces salivation and a reduction in heart rate by a cholinergic mechanism in man. Am J Gastroenterol 1989;84:1147. Richter JE, Castell DO. Drugs, foods and other substances in the cause and treatment of reflux esophagitis. Med Clin North Am 1981 ;65:1223-34. Deveney CW, Benner K, Cohen J. Gastroesophageal reflux and laryngeal disease. Arch Surg 1993;128:1021-7. Kamel P, Kahrilas PJ, Hanson DG, McMahan J. Prospective trial of omeprazole in the treatment of "reflux laryngitis." Gastroenterology 1992;102:A93. Koufman J. Gastroesophageal reflux disease in otolaryngology. Presented at the Annual Meeting of the American Academy of Oto-, Head and Neck Surgery, San Diego, CA, 1994.