Introduction
Over the last 15 years, numerous clinical trials have gradually added evidence that duodenogastro-esophageal reflux reaching the laryngopharynx plays an important aetiological role in the pathogenesis of laryngeal, pharyngeal and pulmonary disorders. While patients suffering from gastroesophageal reflux disease usually complain of heartburn, chest pain and acid regurgitation, these symptoms are rarely found in patients with laryngopharyngeal reflux (LPR). The typical so-called supra-oesophageal reflux symptoms are globus pharyngeus, frequent throat-clearing, chronic cough, hoarseness, asthma, dysphagia and sore throat.Reference Koufman, Aviv, Casiano and Shaw1
Although oesophagogastroduodenoscopy with biopsy is not the method of choice for diagnosis of LPR, this diagnostic approach might be indicated in some patients with pH-documented LPR, in order to exclude Barrett's metaplasia, Helicobacter pylori- associated chronic gastritis, or other gastric or oesophageal abnormalities. Helicobacter pylori- associated chronic gastritis is involved in the development of duodenal and gastric ulcer, gastric adenocarcinoma, and mucosa-associated lymphoid tissue. ‘Barrett's oesophagus’ is a precancerous condition in the development of oesophageal adenocarcinoma. The incidence of oesophageal adenocarcinoma has increased steadily over the last decades at a faster rate than any other type of cancer.Reference Labenz2, Reference Van Soest, Dieleman, Siersema, Sturkenboom and Kuipers3
Another argument for routine oesophagogastroduodenoscopy in LPR patients could be provided by Reavis and colleagues' results demonstrating that LPR symptoms were better indicators of oesophageal adenocarcinoma than were gastroesophageal symptoms.Reference Reavis, Morris, Gopal, Hunter and Jobe4
Currently, few studies have analysed the prevalence of Barrett's oesophagus or H pylori-associated chronic gastritis in patients with LPR. Therefore, the goal of this retrospective study was to determine the rates of Barrett's metaplasia, H pylori-associated chronic gastritis, and other pathological gastric and oesophageal findings, in consecutive patients with pH-documented LPR diagnosed by oesophagogastroduodenoscopy and biopsy.
Patients and methods
Between January 2005 and December 2006, 28 consecutive patients (15 women, 13 men; mean age 50 years) had presented to the Department of Otorhinolaryngology, Head and Neck Surgery at Ludwig Maximilians University Munich, with nonspecific otolaryngological symptoms (Table I). All patients had been suspected of having LPR, based on their symptoms and laryngeal findings, using the reflux symptom index and the reflux finding score.Reference Belafsky, Postma and Koufman5, Reference Belafsky, Postma and Koufman6 None of the patients had been treated with proton pump inhibitors (PPIs) or any other anti-reflux medication at the time of presentation.
Patients presenting symptoms
In order to confirm the diagnosis of LPR, all patients had undergone 24-hour pH-monitoring, using a pH device consisting of one proximal and one distal probe with a fixed distance of approximately 10 cm between the probes. After application of a local, decongestant anaesthetic spray, the recording probe of the device had been placed transnasally. The proximal probe had been positioned immediately above the upper oesophageal sphincter, under the guidance of flexible or rigid laryngoscopy. Patients had kept a diary, recording times and durations of meals. No oesophageal manometry had been performed.
To record the data, we used the pH response reflux diagnostic system (Medtronic Xomed, Jacksonville, Florida, USA). All meal intervals and the first two minutes of the immediate postprandial period had been eliminated from analysis. In order to exclude artefacts from analysis, upper probe events with pH<4 had only been accepted as proximal reflux episodes when an association with a lower probe event could be clearly identified. Such pH-monitoring enabled determination of the reflux area index, currently the most useful parameter for measuring laryngopharyngeal reflux severity.Reference Vincent, Garrett, Radionoff, Reussner and Stasney7 This parameter not only incorporates the number and duration of proximal reflux events, but also the degree to which these episodes drop below pH 4. The reflux area index is a measure of the area subjected to pH<4, corrected for the duration of the study for each patient.Reference Vincent, Garrett, Radionoff, Reussner and Stasney7 Laryngopharyngeal reflux had been diagnosed when the reflux area index was >6.3 for the total pH study duration.
In 10 patients, oesophagogastroduodenoscopy with biopsy had been performed by gastroenterologists in private practice, within a maximum period of six weeks before presentation to our department. Four patients had been referred to our department by the gastrointestinal function laboratory of the department of surgery of Ludwig Maximilians University, due to abnormal laryngeal findings diagnosed during oesophagogastroduodenoscopy (Figure 1). Due to suspected LPR, 14 patients had first undergone 24-hour pH-monitoring in our department and had then been referred to the gastrointestinal function laboratory of our University's surgical department, in order to exclude additional gastroesophageal reflux disease, conditions associated with gastroesophageal reflux disease, or H pylori-associated chronic gastritis. All oesophagogastroduodenoscopy biopsies had been examined by either pathologists of the Ludwig Maximilians University pathology department or by pathologists in private practice (n=10).
Larynx of patient 10 (see Table II), suffering from combined pH-documented laryngopharyngeal reflux (reflux area index >6.3) and gastroesophageal reflux disease. Oesophagogastroduodenoscopy revealed grade B oesophagitis and a large, reflux-induced granuloma of the larynx.
Results
The reflux symptom index had been greater than 13 and the reflux finding score greater than seven in all 28 patients, strongly indicating LPR.Reference Belafsky, Postma and Koufman5, Reference Belafsky, Postma and Koufman6 Additional 24-hour pH-monitoring had also revealed a reflux area index of >6.3 in all 28 patients, reflecting abnormal proximal reflux.Reference Vincent, Garrett, Radionoff, Reussner and Stasney7
No oesophagogastroduodenoscopy abnormalities had been detected in 13 patients (46 per cent). Oesophagitis, classified according to the Los Angeles classification, had been diagnosed in five patients (grade A in four patients, grade B in one patient). Hiatus hernia had been detected in 10 patients (36 per cent) and gastric mucosal erosions in seven patients (25 per cent).
Histological examination of biopsies had revealed H pylori-associated chronic gastritis in two patients (7 per cent) and Barrett's metaplasia in two patients. In one patient, both Barrett's metaplasia and H pylori-associated chronic gastritis had been diagnosed (patient two, Table II). Oesophagitis grade B or Barrett's metaplasia had only been found in patients with heartburn as their main presenting symptom. Data for each patient are summarised in Table II.
Patient data
Pt patient; OGD oesophagogastroduodenoscopy; NA no abnormality; HPG Helicobacter pylori-associated chronic gastritis; O1 oesophagitis grade A; HH hiatus hernia; BM Barrett's metaplasia; GME gastric mucosal erosions; O2 oesophagitis grade B
Discussion
Gastroesophageal reflux disease is considered the most prevalent gastrointestinal disease in Western countries. It is estimated that in central and northern Europe, between 18 and 40 per cent of the population suffer from gastroesophageal reflux disease. This frequent condition is characterised by various clinical manifestations of abnormal reflux of gastric contents from the stomach into the oesophagus.Reference El-Serag and Sonnenberg8, Reference DeVault and Castell9 The most common clinical manifestations of gastroesophageal reflux disease are heartburn and acid regurgitation. When stomach content flows back up to the larynx or even to higher regions of the head and neck, it can cause a wide range of symptoms, commonly termed laryngopharyngeal reflux.
The incidence of heartburn in Laryngopharyngeal reflux patients, however, is less than 40 per cent.Reference Koufman, Aviv, Casiano and Shaw1 In our study, only five patients (18 per cent) had been reported as suffering from heartburn. Nevertheless, in approximately 50 per cent of LPR patients, abnormal gastroesophageal reflux can be identified by pH study, potentially causing such complications as oesophagitis, strictures, Barrett's oesophagus and oesophageal adenocarcinoma.Reference Byrne, Power, Lawlor, Ravi and Reynolds10
Since the 1980s, the incidence of oesophageal adenocarcinoma has been rising by 4–10 per cent annually, and this increasing trend is still continuing.Reference Van Soest, Dieleman, Siersema, Sturkenboom and Kuipers3 Barrett's oesophagus is the most recognised risk factor for oesophageal adenocarcinoma, and the main risk factor for Barrett's oesophagus is gastroesophageal reflux disease.Reference Van Soest, Dieleman, Siersema, Sturkenboom and Kuipers3 The prevalence of Barrett's oesophagus among gastroesophageal reflux disease patients is estimated to range between 5 and 20 per cent.Reference Halum, Postma, Bates and Koufman11 The incidence of oesophageal adenocarcinoma in patients with Barrett's metaplasia is estimated to be 0.5 per cent per year.Reference Coppola and Karl12
Therefore, one could argue that oesophagogastroduodenoscopy should be performed as a routine procedure in patients with pH-documented LPR, in order to rule out additional Barrett's metaplasia. However, little is known about the prevalence of Barrett's oesophagus or oesophagitis in LPR patients. In a recently published study by Halum et al., the prevalence of Barrett's metaplasia in LPR patients was 3 per cent.Reference Halum, Postma, Bates and Koufman11 One study by Koufman et al. found oesophagitis in 12 per cent and Barrett's metaplasia in 7 per cent of patients with LPR.Reference Koufman, Belafsky, Bach, Daniel and Postma13 These findings are similar to our results, with a prevalence of oesophagitis in 18 per cent and Barrett's metaplasia in 7 per cent of patients with pH-documented LPR. As Barrett's metaplasia might only be prevalent in 3 to 7 per cent of LPR patients, it does not, in our opinion, represent an indication to perform routine oesophagogastroduodenoscopy in such patients. However, in those LPR patients with heartburn as their main symptom, oesophagogastroduodenoscopy might be indicated; in our study, Barrett's metaplasia or oesophagitis grade B had only been found in patients complaining mainly of heartburn (i.e. patients 2, 8 and 10; Table II).
A second reason for performing oesophagogastroduodenoscopy as a routine diagnostic procedure in patients with pH-documented LPR could be to rule out H pylori colonisation, this being a strong risk factor for peptic ulceration and distal gastric cancer.Reference Peek14 Moreover, long-term PPI therapy for LPR in the presence of H pylori-associated chronic gastritis may accelerate the loss of specialised glands, leading to atrophic gastritis and potentially to gastric cancer.Reference Kuipers, Nelis, Klinkenberg-Knol, Snel, Goldfain and Kolkman15, Reference Malfertheiner, Megraud, O'Morain and Bazzoli16
Laryngopharyngeal reflux management recommendations call for initial treatment with twice-daily PPIs for a minimum period of six months, with many patients requiring prolonged chronic treatment; H pylori eradication might therefore be especially indicated in many LPR patients.Reference Koufman, Aviv, Casiano and Shaw1, Reference Vaezi17
To our knowledge, there is only one published study investigating the role of gastric H pylori infection in LPR disease. The results of this study, by Ercan et al., demonstrated no relationship between H pylori-associated chronic gastritis and LPR.Reference Ercan, Cakir, Uzel, Sakiz, Karaca and Turgut18 However, according to Oridate et al., H pylori seropositivity predicts the outcomes of PPI therapy for LPR symptoms.Reference Oridate, Takeda and Yamamoto19 Little is known about the prevalence of H pylori-associated chronic gastritis in LPR patients. In our study, we found H pylori-associated chronic gastritis in two patients (7 per cent) from Munich, Germany, whereas Ercan et al. reported H pylori-associated chronic gastritis to be present in 22 of 32 LPR patients (69 per cent) living in Turkey.Reference Ercan, Cakir, Uzel, Sakiz, Karaca and Turgut18 These different results could be explained by the fact that certain ethnic groups are more prone to develop H pylori-associated chronic gastritis than others. In a study by Loffeld et al., H pylori was present in 60.6 per cent of Turkish patients living in the Netherlands but in only 18.5 per cent of Dutch patients.Reference Loffeld20
As mentioned above, H pylori testing should be considered in LPR patients who anticipate requiring long-term maintenance therapy with PPIs. However, oesophagogastroduodenoscopy is not the method of choice for the diagnosis of H pylori-associated chronic gastritis, as non-invasive tests with a very high degree of diagnostic accuracy exist (such as the C-urea breath test, stool antigen tests and various immunological tests).Reference Malfertheiner, Megraud, O'Morain and Bazzoli16 Therefore, oesophagogastroduodenoscopy is not indicated as a routine diagnostic procedure in LPR patients to exclude H pylori-associated chronic gastritis. It remains unclear which LPR patients would benefit from routine H pylori testing before initiation of PPI therapy.
In our study patients, we found a relatively high prevalence of hiatus hernia (36 per cent), a protrusion of the upper part of the stomach into the thorax through a tear or weakness in the diaphragm. Older trials have found hiatus hernia in between 14 and 22 per cent of subjects undergoing upper endoscopy.Reference Wright and Hurwitz21, Reference Cronstedt, Carling, Vestergaard and Berglund22 As these trials, with large study populations ranging from 293 to 1000 subjects, did not specifically focus on LPR patients, the different prevalence data are difficult to interpret. The frequency of hiatus hernia in asymptomatic individuals would be of great interest, but reliable data are unavailable as most studies relate to patients suffering from gastroesophageal reflux disease symptoms undergoing investigation.
As hiatus hernia impairs the lower oesophageal sphincter function by reducing its length and pressure, oesophagitis or Barrett's oesophagus are common sequels of reflux in subjects with hiatus hernia.Reference Gordon, Kang, Neild and Maxwell23 Cameron found hiatus hernia in 71 per cent of patients with oesophagitis and in 72 per cent of patients with short segment (<2 cm) Barrett's oesophagus.Reference Cameron24 Currently, there is no adequate, published evaluation of the frequency of hiatus hernia in LPR patients. Should future trials confirm a high prevalence of hiatus hernia in LPR patients, routine oesophagogastroduodenoscopy might be discussed in these patients in order to rule out this predisposing factor for Barrett's oesophagus. Alternative diagnostic procedures would comprise barium studies or oesophageal manometry.Reference Gordon, Kang, Neild and Maxwell23 However, the role of hiatus hernia in the pathogenesis and progression of gastroesophageal reflux disease and its complications remains unclear.Reference Gordon, Kang, Neild and Maxwell23 Therefore, at present, there is no indication for hiatus hernia screening of LPR patients by routine oesophagogastroduodenoscopy.
Gastric mucosal erosions were another frequent finding in our study population, with a prevalence of 25 per cent. These are common in both dyspeptic patients and asymptomatic volunteers.Reference Lehmann, Renner, Meyer-Wyss, Wilder-Smith, Mazzucchelli and Ruchiti25 A long-term study by Toljamo et al., with a 17-year follow up, showed that a significant proportion of gastric mucosal erosions were chronic or recurrent but were generally without major systemic or local consequences.Reference Toljamo, Niemelä, Karttunen, Karvonen and Lehtola26 Therefore, oesophagogastroduodenoscopy is not generally indicated in LPR patients to rule out gastric mucosal erosions, except in H pylori positive patients, as the latter are at a significantly greater risk of developing peptic ulcers.Reference Toljamo, Niemelä, Karttunen, Karvonen and Lehtola26
• Recently, numerous clinical trials have added evidence that duodenogastroesophageal reflux reaching the laryngopharynx plays an important role in the pathogenesis of laryngeal, pharyngeal and pulmonary disorders
• Patients with laryngopharyngeal reflux (LPR) uncommonly suffer from conditions associated with gastroesophageal reflux disease, such as oesophagitis or oesophageal neoplasms; however, in a small number of such patients, oesophagitis and Barrett's metaplasia occur
• Symptoms such as heartburn or chest pain are indicators of additional gastroesophageal reflux disease in patients with pH-documented LPR; in these patients, endoscopy is indicated
• Helicobacter pylori testing, by non-invasive tests such as the C-urea breath test, should be considered in LPR patients requiring long-term proton pump inhibitor treatment
The results of our study confirm the findings of other trials reporting a prevalence of oesophagitis in LPR patients of less than 20 per cent.Reference Koufman, Aviv, Casiano and Shaw1, Reference Koufman, Belafsky, Bach, Daniel and Postma13 We found only mild forms of oesophagitis (grades A and B, Los Angeles classification), in 18 per cent of the study population. Hence, there is no indication for routine oesophagogastroduodenoscopy in LPR patients in order to rule out oesophagitis as a potential risk factor for oesophageal adenocarcinoma. Moreover, the generally recommended PPI dosage for LPR treatment (i.e. twice daily for three to six months) is much higher than that for mild oesophagitis. In consequence, oesophagitis in LPR patients should be cured by the long-term PPI therapy in most cases. An exception would be patients in whom PPIs were partially or entirely ineffective. In gastroesophageal reflux disease patients, the prevalence of so-called PPI non-responders ranges from 10 to 30 per cent; however, a PPI failure rate of up to 40 per cent has been found in LPR patients.Reference Galmiche and Stephenson27, Reference Vaezi28
Conclusions
From our findings, we conclude that not all patients with pH-documented LPR should routinely undergo oesophagogastroduodenoscopy. However, in patients with heartburn as their main symptom, Barrett's metaplasia or severe forms of oesophagitis should be excluded by oesophagogastroduodenoscopy. Non-invasive H pylori testing (e.g. the C-urea breath test) should be considered in LPR patients requiring long-term PPI treatment. In these patients, H pylori eradication might be indicated before PPI therapy. If future studies clearly demonstrate a strong influence of hiatus hernia on the pathogenesis and progression of gastroesophageal reflux disease and its complications, then the currently recommended diagnostic and management strategies of LPR may need to be reviewed.
