© Springer International Publishing AG 2018Eytan Bardan and Reza Shaker (eds.)Gastrointestinal Motility Disorders doi.org/10.1007/978-3-319-59352-4_16
16. Dysphonia and Laryngopharyngeal Reflux
Department of Otolaryngology—Head and Neck Surgery, Medical College of Georgia at Augusta University, 1120 15th Street, Suite BP 4109, Augusta, GA 30912, USA
KeywordsRefluxLaryngopharyngeal refluxExtraesophageal refluxDysphoniaHoarsenessStroboscopyLaryngovideostroboscopy
“My doctor says my reflux is causing my hoarseness. What do you think?”
Sally comes to the office today with a lot of nonspecific symptoms including throat clearing and vocal fatigue, and is concerned that her voice has become hoarse and she cannot sing in the choir anymore. She has already seen her primary care provider who attributed the problem to acid reflux. She began her proton pump inhibitor medication 10 weeks ago, and while she has not taken it daily as directed, she feels that there might be a “slight improvement.” She then presents to your office with the above question and is obviously still very concerned about not being able to sing in the choir anymore.
The above story is a very typical scenario for an otolaryngologist or a gastroenterologist. Prior to the publication of Koufman’s landmark thesis on laryngopharyngeal reflux (LPR) in 1991 , the very possibility of extraesophageal reflux was nearly always ignored but the reverse became true over the next 15 years. LPR became the cause of literally any symptom in the head and neck often without any examination of the laryngopharynx. Over the past few years this has been appropriately questioned and we would suggest that a “middle ground” appears reasonable. So for our patient Sally, our answer would be that let us consider other possible causes of your symptoms and get started by examining your larynx and pharynx to appreciate any findings that can guide further evaluation and treatment. As we will discuss, not everything in the throat or a change in voice can be attributed to acid reflux and it is the role of the clinician to tease out the other causes before embarking on long-term therapy with a proton pump inhibitor or surgical treatment of reflux.
LPR has been blamed for a multitude of symptoms over the years, including hoarseness, chronic cough, excessive throat clearing, vocal fatigue, postnasal drip, globus pharyngeus, and dysphagia . Furthermore, it has also been blamed at least in part for the pathogenesis of multiple lesions in the larynx including vocal process granulomas, subglottic stenosis, muscle tension dysphonia, laryngospasm, and even laryngeal carcinoma . Since the early 1990s, there was a move toward attributing much of what we did not know or could not visualize in the larynx as being due to LPR. In many ways, it became the default diagnosis for much of the above problems and symptoms. Proton pump inhibitors provided an easy enough empiric therapy before the patient could see the specialist, but even with visualization of the larynx by some otolaryngologists it became the “waste basket” diagnosis when there was nothing else grossly evident on indirect visualization with a mirror or flexible laryngoscopy in the office.
Over the past several years, there has been a push toward better visualization and need to exclude other pathologies such as benign vocal fold lesions that would benefit from surgical excision, muscle tension dysphonia that would be treated with speech therapy, vocal fold paresis that could benefit from a vocal fold injection to assist glottal closure, and so on. Before launching into empiric therapy for LPR, it behooves the clinician to look deeper for the causes of laryngeal symptoms in order to arrive at appropriate therapy.
Voice disorders have a lifetime prevalence of around 30% . Reflux has been previously implicated or at least associated with a large number of these voice disorders in the last few decades in multiple clinical studies [5–10]. Koufman et al. described 113 unselected, new patients in an often-cited study from 2000 that presented with a variety of laryngeal and voice disorders . After extensive history and physical examination was performed, all patients with signs and symptoms of LPR (78/113) were sent for ambulatory 24-h double-probe pH monitoring. The authors defined abnormal as a single hypopharyngeal event of pH below 4.0 immediately preceded by a similar drop in the esophageal pH probe, which produced a large number of tests as being abnormal at 57/78 or 50% of the total number that presented. Of these patients, those muscle tension dysphonia had 70% with a concomitant LPR diagnosis compared to neuromuscular conditions with a much lower 19% LPR association.
Ozturk et al. in 2006 attempted to find the prevalence of LPR in those patients with hoarseness and compare it to normal controls . They enrolled 43 patients with hoarseness of greater than 3 months and recruited 20 healthy volunteers. All subjects then received flexible larygngoscopic evaluation as well as 24-h double-probe pH monitoring. 62.8% of the study group had LPR episodes compared to 30% of the healthy population. While there was increased incidence of LPR in the study population, they also showed increased severity of LPR in the study population with significantly increased mean total number of LPR episodes. The most common symptoms of their study population were heartburn and chronic throat clearing and their most common physical finding was a very nonspecific finding of posterior laryngeal pachydermia which was seen in 67% of patients. However, they do not quantitate how many of these patients have LPR on pH testing. They do break down the data to show that 61.8% of those patients with evidence of posterior laryngitis on flexible laryngoscopy had LPR on testing, and 72.7% of those patients with vocal fold lesions (including polyps, nodules, granulomas, leukoplakia, and subglottic stenosis) did as well. Their study is able to show an association between LPR and endoscopic findings but lacks in drawing clear causations and showing data on how many people with these endoscopic findings have LPR.
In a different design, Cohen and Garrett aimed to look at the prevalence of PPI therapy in those patients referred with hoarseness and compare it with their final diagnostic findings . They performed a retrospective review of 299 such patients that were either taking or had taken PPIs in the previous 2 months prior to referral to a tertiary care voice clinic. They found an almost 3:1 female-to-male ratio of patients that met their inclusion criteria, and among those referred, they found that 56.1% had been on or were currently on PPI medications. 29.7% had stopped taking their medication due to continued hoarseness and the other 70.3% had persistent hoarseness despite PPI therapy. They concluded that it was extremely common for patients referred to a tertiary care voice clinic to already have undergone PPI therapy for treatment of their hoarseness but it was of questionable benefit. Of those that quit taking their PPIs, 79.5% of them did not have traditional GERD symptoms of heartburn or regurgitation. The four most common diagnoses after the patient was seen were muscle tension dysphonia , benign vocal fold lesion, GERD, and vocal fold paralysis. Their study showed how common it is for patients to be started on PPI therapy without evidence of LPR leading to delayed diagnoses and persistent symptoms, and contributing to rising healthcare costs.
Karkos and colleagues attempted to look at a cohort of patients with persistent functional dysphonia and see if there were any pH testing findings that distinguished them from controls . They took 23 patients seen in a voice clinic who had persistent functional dysphonia (defined as lack of any major structural or neurologic abnormality by stroboscopy) for 3 months and compared their 24-h dual-probe pH-metry to 8 healthy controls. They found that most pH-metry parameters did not show any statistical differences between the patients and controls. While they did show this association between certain pH testing findings and functional dysphonia, they could not conclude any causation.
Patel, Carroll, and colleagues attempted to look at the pathology of vocal fold atrophy and determine whether their patients’ symptoms of throat clearing and mucus sensation that had been attributed to LPR were really from glottic insufficiency . They retrospectively looked at 26 patients with vocal fold atrophy that had these symptoms and that had been initially attributed to LPR. They found that the majority of their patients improved their reflux symptom index (RSI) scores with a mix of fold augmentation and voice therapy. Specifically though, those patients that had glottic insufficiency that had injection augmentation had improvement in their RSI and voice after treatment. In conclusion, they argued that these symptoms should not only be due to LPR but may also be due to underlying vocal fold atrophy and resulting glottic insufficiency.
Gastroesophageal reflux (GERD) and LPR are not the same disease processes and should not be used interchangeably, but studies have shown a direct correlation between patients’ severity of GERD as found by UGI endoscopy and symptom-proven GERD and the prevalence of LPR by use of a questionnaire . Taken a different way, researchers have also described the prevalence of GERD in those patients presenting with only laryngeal complaints . Researchers looked at 30 patients presenting to the otolaryngologist and were then all sent for esophagogastroduodenoscopy (EGD) as part of the study. The patients had a variety of complaints ranging from dysphagia, globus pharyngeus, hoarseness, odynophagia, sore throat, excessive mucus production, throat clearing, laryngospasm, and vocal fatigue. The researchers found a large prevalence of gastroenterological diseases including GERD (43%), hiatal hernia (43%), and Helicobacter pylori-positive antrum gastritis (23%). Interestingly, treatment with antireflux medications and eradication of Helicobacter pylori infection were able to completely resolve laryngopharyngeal symptoms and findings in 20 of 22 patients (90% success rate) during their 8-month follow-up period. Additionally, they found medical antireflux medication to be effective in the treatment of laryngopharyngeal symptoms.
All of these clinical studies contributed to the association between hoarseness and other laryngeal symptoms with laryngopharyngeal reflux. As the reader can tell, there is a lack of definitions of what constitutes laryngopharyngeal reflux by endoscopic findings as well as even what constitutes an abnormal dual-probe pH testing making drawing clear conclusions difficult for the clinical researcher and even harder for the clinician in the real world.
Basic Science Studies
There have been multiple basic science studies performed in the last 50 years trying to decipher mechanisms of injury from reflux in the larynx. A study in 1968 by Delahunty and Cherry was the first to show in an animal model how gastric acid could cause laryngeal injury. They described a model for vocal fold granuloma development by exposing canine larynges to gastric acid . In their study, they secured gastric acid refluxate from the dogs via nasogastric tube. They then soaked a cotton ball in the refluxate, and placed it on the posterior half of the vocal fold in two dogs every day for 4 weeks. After 4 days, they were able to show mucosal changes, and after 4 weeks, the affected vocal cord of each canine was markedly inflamed, thickened, and irregular. At day 29 the vocal fold started sloughing, leading to formation of friable granulation tissue that replaced the vocal fold epithelium. Obviously, patients do not have such a high delivery of their own gastric acid to their larynx, but it was the first to show what could happen with chronic acid exposure nonetheless.
In terms of the actual mechanism of injury of LPR on the larynx, an international group of researchers looked at the effect of LPR on laryngeal epithelium to examine the impact of the disease at the cellular level . They looked at carbonic anhydrase, E-cadherin, and MUC gene expression in LPR patients, controls, and an in vitro model. The LPR patients had a decreased carbonic anhydrase level in the vocal fold epithelium but increased levels in the posterior commissure epithelium. Their studies showed a depletion of carbonic anhydrase in reaction to reflux events. In addition, E-cadherin was not even present in 37% of LPR specimens obtained, implying that the larynx does not have a key transmembrane cell surface molecule that can help in epithelial defense making it more susceptible to injury from reflux. Other studies in their group showed that there was indeed injury from reflux at pH above 4 which is classically the pH level under which pepsin becomes activated . In a separate article, pepsin was shown to be co-localized with transferrin in intracellular vesicles, implying that pepsin can be taken up into laryngeal epithelial cells by receptor-mediated endocytosis. This pathway of pepsin into the cells established at least a theoretical mechanism for pepsin damage to the larynx.
Clinical Studies Revisited
More recently, there has been an emphasis from the laryngology community toward eliminating the “waste basket” diagnosis of LPR to the detriment of other causes that then contributes to putting patients on unnecessary medications and delaying the diagnosis and treatment of real underlying pathologies. Patients with a voice complaint will likely present to their primary care physician initially for care. Ruiz et al. surveyed these primary care providers in their treatment of dysphonia , with special attention to LPR . 12.9% of the surveyed physicians responded, yielding 314 completed surveys. The authors found from the survey that most of these physicians preferred to treat patients with hoarseness before referring to an otolaryngologist. Reflux medications (85.8%) and antihistamines (54.2%) were the most commonly used medications for this empirical treatment. 79.2% of these physicians would also treat chronic hoarseness with reflux medication in a patient without evidence or symptoms of gastroesophageal reflux disease. The authors went on to conclude that primary care physicians often attribute dysphonia to reflux. However, they did not look at any patients in particular and just looked at the very high prevalence of PPI therapy without any visualization of the larynx.