glands. Oral bacteria chemically transforms the relatively innocuous nitrate to the more toxic nitrite,

which is swallowed and subsequently converted to nitric oxide and other toxic nitroso-compounds by

acid and ascorbic acid in the stomach. Whether this mechanism in fact contributes to injury and/or

neoplastic transformation in the upper stomach, gastroesophageal junction, and distal esophagus is

currently unknown.

Figure 42-30. Peak bile acid concentration (μmol/L) for patients and normal subjects during upright, postprandial, and supine

aspiration periods. The shaded area represents the mean and the bar the 95th percentile values.

Respiratory Complications

It is increasingly recognized that a significant proportion of patients with gastroesophageal reflux will

have either primary RSs or RSs in association with more prominent heartburn and regurgitation.94

Reflux has been implicated as causative of asthma and idiopathic pulmonary fibrosis (IPF) and can

complicate advanced lung diseases including chronic obstructive pulmonary disease (COPD) and cystic

fibrosis (CF). Thirty-five to fifty percent of asthmatics have been shown to have abnormal esophageal

pH, esophagitis, and a hiatal hernia.95 Others have shown that the prevalence of reflux symptoms

exceeded 50% in patients with asthma and CF.96,97 In addition, patients with COPD and

gastroesophageal reflux are twice as likely to have significant COPD exacerbations than their nonreflux

counterparts,98 and reflux symptoms correlate with airway obstruction in COPD patients. These reports

suggest that the frequency of dual pathology is higher than would be expected by chance alone.

Pathophysiology of Reflux-Induced Respiratory Symptoms. Two mechanisms have been proposed

as the pathogenesis of reflux-induced RSs. The first, the so-called “reflux” theory, maintains that RSs are

the result of the aspiration or microaspiration of gastric contents. The second or “reflex” theory

maintains that vagally mediated bronchoconstriction follows acidification of the lower esophagus via a

reflex neurologic arc.

The evidence supporting a reflux mechanism was shown in animals in the 1950s to 1970s, with studies

of acid instillation in the lungs resulting in pneumonitis,99,100 epithelial damage,101 and increased airway

resistance.102 The evidence in humans can be categorized as follows. First, there is radiographic

evidence. Patients with pulmonary fibrosis have an increased prevalence of hiatal hernia on upper GI

series,103,104 while elderly patients with large hiatal hernias have a significantly increased frequency of

pleural scarring on chest CT scans.105 Second, erosive esophagitis was found at a high frequency in a

large series of Veterans Affairs patients with asthma, pneumonia, chronic bronchitis, and pulmonary

fibrosis.94 Third, patients with chronic lung diseases, including asthma and pulmonary fibrosis in

particular, have a significantly higher frequency of positive 24-hour pH studies.105,106 With dual pH

monitoring of the proximal and distal esophagus, prolonged exposure of the proximal esophagus to

gastric juice occurs at a higher frequency in patients with gastroesophageal reflux and RSs than in

patients without respiratory complaints.107,108 Fourth, simultaneous tracheal and esophageal pH

monitoring in patients with reflux disease have documented tracheal acidification in concert with

esophageal acidification.102 Finally, medications used to treat chronic lung diseases, in particular,

bronchodilators, relax esophageal sphincter tone.109

A reflex mechanism is primarily supported by the fact that bronchoconstriction occurs following the

infusion of acid into the lower esophagus.110–112 This can be explained by the common embryologic

1069

origin of the tracheoesophageal tract and its shared vagal innervation.

The primary challenge in implementing treatment for reflux-associated RSs lies in establishing the

diagnosis. In those patients with predominantly typical reflux symptoms and secondary respiratory

complaints, the diagnosis may be straightforward. However, in a substantial number of patients with

reflux-induced RSs, the RSs dominate the clinical scenario. Gastroesophageal reflux in these patients is

often silent and is only uncovered when investigation is initiated.106,113 A high index of suspicion is

required, notably in patients with poorly controlled asthma in spite of appropriate bronchodilator

therapy. Supportive evidence for the diagnosis can be gleaned from endoscopy and stationary

esophageal manometry. Endoscopy may show erosive esophagitis or Barrett esophagus. Manometry

may indicate a hypotensive lower esophageal sphincter or ineffective body motility, defined by 30% or

more contractions in the distal esophagus of less than 30 mm Hg in amplitude.114 D’Ovidio et al.115

studied 78 patients awaiting lung transplant for IPF, scleroderma, COPD, or CF and found that 72% had

a hypotensive lower esophageal sphincter, 33% abnormal esophageal body motility, and 38% abnormal

24-hour pH testing, although, interestingly, patients with IPF in other series have been shown to have

normal manometry.108

The current “gold standard” for the diagnosis of reflux-induced respiratory complaints is ambulatory

dual-probe pH monitoring, often combined with multichannel intraluminal impedance (MII-pH). One

probe is positioned in the distal esophagus and the other at a more proximal location. Sites for proximal

probe placement have included the trachea, pharynx, and proximal esophagus. Most authorities would

agree that the proximal esophagus is the preferred site for proximal probe placement. While ambulatory

esophageal pH monitoring allows a direct correlation between esophageal acidification and RSs, the

chronologic relationship between reflux events and bronchoconstriction is complex. Further, the

proximal probe must be positioned close to, but not above, the UES. Failure to properly place the probe

can lead to artifact or errors in pH assessment. In addition, the use of a pH threshold of 4 as is used in

the distal esophagus is questioned. Normal subjects should have a pH ≥7 in the cervical esophagus for

at least 19.6% of the monitored period. Looking at pH exposure in reverse, in other words as loss of the

normal alkalinity in the cervical esophagus, was shown to be a more sensitive indicator of proximal

reflux.116 Recently the use of a special pH probe designed for the environment of the pharynx has been

advocated in patients with RSs to evaluate for proximal reflux. This probe, called the Restech pH probe

(Respiratory Technology Corp, San Diego, CA, USA), has been shown to better predict successful relief

of extraesophageal RSs with a fundoplication compared to the traditional dual-probe pH test.117 Of

importance, the Restech analysis uses a pH threshold of 5.5 for upright and 5.0 for supine reflux, and

the Ryan score is used to determine normal from abnormal pharyngeal reflux exposure.116 Probably the

best pH monitoring method in these patients is to start with a 48-hour Bravo pH and if that is normal on

both days then add a Restech pH test.117

Laryngopharyngeal Reflux

The term LPR is used for a broad category of atypical symptoms that may be secondary to

gastroesophageal reflux. These may include laryngitis, pharyngitis, hoarseness, sinusitis, sleep

disturbance, dental erosions, and globus.118,119 Mechanisms for these symptoms are difficult to prove

and substantial evidence remains lacking.120

Treatment of Extraesophageal Complications. Once gastroesophageal reflux is suspected or thought

to be responsible for extraesophageal symptoms, treatment may be with either prolonged PPI therapy

or antireflux surgery. Initially, a 3- to 6-month trial of high-dose PPI therapy (b.i.d. or t.i.d. dosing)

may help confirm (by virtue of symptom resolution) the fact that reflux is partly or completely

responsible for the symptoms. If symptoms are improved, patients can be treated with maintenance

medical therapy or may be offered surgical treatment to control the reflux. The persistence of symptoms

despite PPI treatment, however, does not necessarily rule out reflux as being a potential contributor.

Algorithm 42-1 bases clinical decisions on the results of dual-probe 24-hour pH monitoring and is a

useful starting point.

1070

Algorithm 42-1. 24-Hour Esophageal pH Monitoring

Based on reported observations, relief of RSs can be anticipated for 25% to 50% of patients with

reflux-induced asthma treated with antisecretory medications,121–123 although patients with pulmonary

fibrosis in particular may have an inherent resistance to standard-dose PPIs.124 Fewer than 15% of

asthmatics with symptom relief, however, can be expected to have objective improvements in their

pulmonary function. The reason for this apparent paradox may be that most studies employed relatively

short courses of antisecretory therapy (<3 months). This time period may have been sufficient for

symptomatic improvement but insufficient for recovery of pulmonary function. The chances of success

with medical treatment are likely directly related to the extent of reflux elimination. The conflicting

findings of reports of antisecretory therapy may well be due to inadequate control of gastroesophageal

reflux in some studies. The literature indicates that antireflux surgery improves RSs in nearly 90% of

children and 70% of adults with asthma and reflux disease.123,125 Improvements in pulmonary function

were demonstrated in around one-third of patients. Comparison of the results of uncontrolled studies of

each form of therapy and the evidence from the two randomized controlled trials of medical versus

surgical therapy indicate that fundoplication is the most effective therapy for reflux-induced asthma.125

The superiority of the surgical antireflux barrier over medical therapy is probably most noticeable in the

supine posture, which corresponds with the period of acid breakthrough with PPI therapy and is the

time in the circadian cycle when asthma symptoms and peak expiratory flow rates are at their worst.

More convincing data establishing the benefit of antireflux surgery in improving RSs and pulmonary

function tests derives from the lung transplantation literature. Bronchiolitis obliterans syndrome (BOS),

synonymous with chronic rejection of the transplanted lung, is diagnosed by a greater than 20% decline

in pulmonary function tests from posttransplant baseline. Lau et al. showed in the early 2000 that 67%

of 18 patients with BOS had an improvement in their pulmonary function tests after antireflux

surgery.126 More recently, several groups have reported improvements in oxygen requirements and

stabilization of declining pulmonary function tests when pre–lung transplant patients underwent

antireflux surgery.127,128 In addition, pulmonary and other extraesophageal symptoms can improve after

laparoscopic Nissen fundoplication with minimal perioperative morbidity.129

It is also important to realize that, in asthmatic patients with a non–reflux-induced motility

abnormality of the esophageal body, performing an antireflux operation may not prevent the aspiration

of orally regurgitated, swallowed liquid or food. This can result in RSs and airway irritation that may

elicit an asthmatic reaction. This factor may be the explanation why surgical results appear to be better

in children than adults, since disturbance of esophageal body motility is more likely in adult patients.

Metaplastic (Barrett) and Neoplastic (Adenocarcinoma) Complications

The condition whereby the tubular esophagus is lined with columnar epithelium rather than squamous

epithelium was first described by Norman Barrett in 1950 (Fig. 42-31).130 He incorrectly believed it to

be congenital in origin. It is now realized that it is an acquired abnormality, occurring in 7% to 10% of

patients with GERD, and represents the end stage of the natural history of this disease.131 It is also

understood to be distinctly different from the congenital condition in which islands of mature gastric

columnar epithelium are found in the upper half of the esophagus.

The definition of Barrett esophagus has evolved considerably over the past decade.130–133

Traditionally, Barrett esophagus was identified by the presence of any columnar mucosa extending at

1071

least 3 cm into the esophagus. Recent data indicating that specialized intestinal-type epithelium is the

only tissue predisposed to malignant degeneration, coupled with the finding of a similar risk of

malignancy in segments of intestinal metaplasia less than 3 cm long, have resulted in the diagnosis of

Barrett esophagus, given any length of endoscopically visible tissue that is intestinal metaplasia on

histology (Fig. 42-32). Whether to call long segments of columnar mucosa without intestinal metaplasia

Barrett esophagus is unclear. The hallmark of intestinal metaplasia is the presence of goblet cells.

Recent studies have identified a high prevalence of biopsy-proven intestinal metaplasia at the cardia, in

the absence of endoscopic evidence of a CLE. The significance and natural history of this finding remain

unknown. The term Barrett esophagus should currently be used in the setting of an endoscopically visible

segment of intestinal metaplasia of any length or columnar replacement of the esophagus of 3 cm or

more.

Figure 42-31. Endoscopic appearance of Barrett esophagus. Note the pink metaplastic mucosa as opposed to the normal whitish

squamous lining of the esophagus.

Factors predisposing to the development of Barrett esophagus include early-onset GERD, abnormal

lower esophageal sphincter and esophageal body physiology, and mixed reflux of gastric and duodenal

contents into the esophagus.133 Direct measurement of esophageal bilirubin exposure as a marker for

duodenal juice has shown that 58% of the patients with GERD have increased esophageal exposure to

duodenal juice and that this exposure is most dramatically related to Barrett esophagus.134

Pathophysiology of Barrett Metaplasia. Recent studies suggest that the metaplastic process at the

gastroesophageal junction may begin by conversion of distal esophageal squamous mucosa to cardiactype epithelium, heretofore presumed to be a normal finding.132 This is likely due to exposure of the

distal esophagus to excess acid and gastric contents via prolapse of esophageal squamous mucosa into

the gastric environment. This results in inflammatory changes at the gastroesophageal junction and/or a

metaplastic process, both of which may result in the loss of muscle function and a mechanically

defective sphincter allowing free reflux with progressively higher degrees of mucosal injury. Intestinal

metaplasia within the sphincter may result, as in Barrett metaplasia of the esophageal body. This

mechanism is supported by the finding that as the severity of GERD progresses, the length of columnar

lining above the anatomic gastroesophageal junction is increased.

Figure 42-32. Histologic appearance of Barrett esophagus. To the left of the photograph, columnar mucosa with abundant goblet

1072