Role of acid and duodenogastric reflux in esophageal mucosal injury: A review of animal and human studies

GASTROENTEROLOGY1995;108:1897-1907

SPECIAL REPORTS AND REVIEWS Role of Acid and Duodenogastric Reflux in Esophageal Mucosal Injury: A Review of Animal and Human Studies MICHAEL F. VAEZI,* SWARNJIT S I N G H , * and JOEL E. RICHTER t *Division of Gastroenterology, Universityof Alabama at Birmingham, Birmingham, Alabama; and tDepartment of Gastroenterology,The Cleveland Clinic Foundation, Cleveland, Ohio

The role of acid and duodenogastric reflux (DGR) in the development of esophageal mucosal injury has been extensively investigated using both animal and human models. In this report, clinical and experimental data are reviewed. The mechanisms by which gastric and duodenal contents produce esophageal mucosal injury are also discussed. Acid and pepsin are unquestionably important in causing mucosal damage at low pH values in both animal and human models. Animal models suggest synergistic damaging potential for conjugated bile acids and HCI as well as that of unconjugated bile acids and trypsin in more neutral pH values. Human evidence for the involvement of bile and its constituents has been controversial; however, the advent of better technology to detect DGR is beginning to clarify the role of these constituents. The contribution of each methodology in clarifying the extent of involvement of DGR in esophageal mucosal injury is reviewed. Despite some conflicting results, preliminary human studies support the results from the animal data suggesting synergistic damaging effects for both bile and acid in esophageal mucosal injury. The implication of these studies in treating g a s t r o e s o p h a g e a l reflux disease are discussed.

astroesophageal reflux disease (GERD) is a common disorder affecting approximately 1 0 % - 3 0 % of Americans. 1 It represents symptomatic retrograde flow of gastric contents into the esophagus with or without associated histological changes. Most healthy individuals intermittently reflux gastric contents into the esophagus. Such episodes occur primarily in the postprandial period, are short-lived, and rarely cause symptoms or esophageal damage. However, excessive gastroesophageal reflux may produce symptoms, and nearly 50% of these patients have complications of GERD, including esophagitis, strictures, or Barrett's esophagus at endoscopy. 2 The noxious agents responsible for injuring the esophageal mucosa originate from both gastric and duodenal sources. Hydrochloric acid and pepsin are the important gastric contents predisposing to the development of

G

esophageal symptoms and mucosal injury. 3 6 Furthermore, gastric juice may intermix with duodenal contents by transpyloric reflux of bile and pancreatic secretions. Regurgitation of duodenal contents into the stomach is known as duodenogastric reflux (DGR) and occurs in normal individuals, especially postprandially 7-1° and at night. ~t'~2 However, when excessive, it may be associated with gastritis, gastric ulcers, postcholecystectomy syndrome, dyspepsia, and possibly esophagitis. 12-23 The duodenal ingredients implicated in esophageal mucosal injury include conjugated and unconjugated bile acids, trypsin, and lysolecithin. 24-26 The relative importance and contribution of acid and DGR to the development of esophageal mucosal injury has been controversial and the subject of many studies using both animal models 3-6'24-49 and human subjects.50 104 Until recently, the main difficulty has been extrapolating the findings in animal studies to humans. However, recent clinical studies using state-of-the-art methodologies for detecting esophageal reflux of acid and duodenal contents have helped unravel the role of these potentially injurious agents in producing esophagitis. This report will review the important animal and human studies that clarify the role of gastric and duodenal contents in causing esophageal symptoms and mucosal damage. Animal Studies Role of Acid and Pepsin

The primary importance of acid and pepsin in causing esophageal mucosal damage has been irrefutably established by investigations using animal models. Using the canine esophagus, Redo et al. 3 investigated the role of acid alone and in combination with various pepsin concentrations by infusing pepsin at concentrations up Abbreviations used in this paper: DGR, duodenogastricreflux; GERD, gastroesophagealrefluxdisease; NAF,net acid flux. © 1995 by the AmericanGastroenterologicalAssociation

0016-5085/95/$3.00

1898

VAEZI ET AL.

to 10% and at pH values of less than and greater than 2. They reported no esophageal damage with HC1 infusion alone, whereas acid in combination with low concentrations of pepsin at pH < 2 caused the most severe esophagitis. Meanwhile, Goldberg et al. 6 showed esophageal mucosal damage in the intact feline esophagus with either very high concentrations of acid (pH 1.0-1.3) or lower acid concentrations (pH 1.6-2.0) in the presence of pepsin. Using a perfused rabbit esophagus model, Lillemoe et al. 27 confirmed that acid infusion alone did not produce mucosal damage or increase esophageal mucosal permeability as measured by the net flux of H +, K +, glucose, and hemoglobin. However, the addition of pepsin to the infusate in a dose-dependent manner was associated with increased degrees of gross esophageal mucosal injury and changes in esophageal mucosal permeability. Thus, the esophageal mucosa is relatively resistant to reflux of acid alone unless it occurs at very high concentrations (pH 1.0-1.3). On the other hand, the combination of acid and even small concentrations of pepsin resuits in macroscopic as well as microscopic esophageal mucosal injury. Separating the role of pepsin from acid in the production of esophagitis is difficult because the optimum pH for the enzymatic activity of pepsin is below 3. 24

Role of Duodenal Contents The role of duodenal contents (specifically, bile acids and the pancreatic enzyme trypsin) in the development of esophageal mucosal injury is controversial and the subject of many in vitro animal studies. 28-32 Early studies by Cross and Wangensteen28 suggested a role for bile and its constituents (namely, bile acids) in esophageal mucosal damage. Using a dog model with biliary diversion and a jejunal conduit anastomosing directly to the esophagus, Moffat and Berkas 29 showed that canine bile was capable of producing various degrees of erosive esophagitis, thereby confirming earlier studies by Cross and Wangensteen. 2s More recent studies show that esophageal mucosal damage by bile acids is dependent on the conjugation state of the bile acids and the pH of the refluxate. Using net acid flux (NAF) across the esophageal lumen as an index of mucosal injury, Harmon et al. 3° showed that taurine-conjugated bile salts (1-5 mmol/L), taurodeoxycholate and taurocholate (both with pKa of 1.9) increased NAF at pH 2, whereas the unconjugated forms (1-5 mmol/L) increased NAF at pH 7 but not at pH 2. Hence, conjugated bile acids are more injurious to the esophageal mucosa at acidic pH, whereas unconjugated bile acids are more harmful at pH 5-8. Using rabbit esophageal perfusion studies, Kivilaakso et al. 3. confirmed the pH-

GASTROENTEROLOGY Vol. 108, No. 6

dependent damage caused by conjugated and unconjugated bile acids (5 and 20 mmol/L) and additionally showed the injurious effect of trypsin (1000 U/mL) on the esophageal mucosa at pH 7.0. Therefore, they concluded that alkaline reflux esophagitis was caused by both unconjugated bile acids and trypsin at neutral pH values.

Synergism Between Acid and Duodenal Contents Because the reflux of gastroduodenal contents usually occurs intermixed with the acidic contents of the stomach, several investigators have studied the synergistic and inhibitory interactions of HC1 with pepsin, trypsin, and bile acids. 25'27'33'34Lillemoe et al. 27 compared the injurious effects of the various duodenal components on rabbit esophageal mucosa at pH 2. At this acidic pH, trypsin (1000 U/mL) had no effect on net flux of ions across the esophageal mucosa because the enzyme is inactive at pH values below 4. Meanwhile, taurocholate (5 mmol/L) produced no esophageal mucosal damage at a neutral pH, but in an acidic medium (pH 1.2), there was esophageal mucosal disruption as measured by net ion permeability. Similarly, Salo and Kivilaakso25 found that both taurocholate (10 mmol/L) and lysolecithin (2 mg/mL), the latter a normal constituent of duodenal juice formed by pancreatic phospholipase A hydrolysis of lecithin in bile, causes histological damage and alteration of the rabbit esophageal transmucosal potential difference in the presence of HC1, but there was no effect in the absence of HC1. Bile acids, depending on their conjugated states, also have both synergistic and inhibitory interactions with trypsin and pepsin. In perfusion studies of the rabbit esophagus, Salo and Kivilaakso33 found that the unconjugated bile acid cholate (10 mmol/L) significantly increased mucosal damage caused by trypsin (1 mg/mL) at pH 7.0. On the other hand, Lillemoe et al. 34 found that the degree of esophageal mucosal injury and permeability decreased in a dose-dependent manner when increasing concentrations of the conjugated bile acid taurodeoxycholate (2, 5, or 10 mmol/L) were added to pepsin (0.3 mg/mL). Therefore, as shown in Figure 1, there is evidence in the animal model for synergism between HC1 and pepsin as well as HC1 and conjugated bile acids and lysolecithin in causing esophageal mucosal damage. Similarly, unconjugated bile acids seem to augment the damaging effect of trypsin at pH 7. HC1 is inhibitory to the damaging effects of trypsin and unconjugated bile acids, whereas conjugated bile acids decrease the damaging effect of pepsin at acidic pH.

June 1995

DUODENOGASTRIC REFLUX 1899

Figure 1. Schematic representation of the postulated injurious agents responsible for esophageal mucosal damage. Mucosal injury is illustrated by the heavy set boxes representing the epithelial surface.

However, we must exercise caution in extrapolating the in vitro animal data on the synergism between bile acids and HC1 in causing esophageal mucosal injury to humans. Experimental animal studies used bile acid concentrations ranging from 1 to 20 mmol/L, which are higher than the bile acid concentrations found in gastric aspiration experiments in humans (0-1 mmol/L). *2,vS-v9 In the only study using bile acid concentrations of < 1 mmol/L, Kiroff et al. 35 found no difference in rabbit esophageal mucosal injury produced by adding 0.1 mmol/L taurocholate to HC1 than acid perfusion alone. However, further experiments are needed to investigate the synergism between bile acids and HC1 in producing esophageal mucosal injury at bile concentrations more representative of those found in the human stomach (0.1 and 1 mmol/L). Mechanisms

of Injury

The mechanism of esophageal mucosal damage by pepsin and trypsin are clearly related to the proteolytic properties of these enzymes. Both promote detachment of the surface cells from the epithelium, presumably by digesting the intercellular substances and surface structures that contribute to the maintenance of cohesion between cells. 36'3v Each agent causes the most damage at its optimal pH activity range: pH 2 - 3 for pepsin and pH 5 - 8 for trypsin. The mechanism for mucosal damage by HCI is more complicated and depends on a series of events. Based on experimental works in the rabbit esophagus, 38-41 H + impairs cell volume regulation causing cell death by inactivation of the Na+/K+-adenosine triphosphatase pump located in the basolateral cell wall in the stratum spinosum of the mucosa. Inhibition of Na+/K+-adenosine triphosphatase occurs at the same time that an amiloride-

sensitive Na + pump is activated, causing increased entry and accumulation of Na ÷ intracellularly, resulting in excess intracellular volume and subsequent cell death. Snow et al. 42 have proposed an alternative mechanism by which acid-induced esophageal mucosal injury may inhibit normal cell volume regulatory mechanisms. Using isolated rabbit esophageal mucosal basal cells, these investigators found pH-dependent alteration in K ÷ and/ or C1- conductance. The mechanism by which bile acids cause mucosal damage is not fully understood. Studies suggest two hypotheses. The first is that bile acids damage mucosal ceils by their detergent property and solubilization of the mucosal lipid membranes. This theory is supported by studies in gastric mucosa in which bile acid-induced mucosal injury was correlated with the release of phospholipids and cholesterol into the lumen. 43-45 However, studies with rabbit esophageal m u c o s a 46'47 show significant mucosal barrier disruption occurring at bile acid concentrations below the level at which phospholipids are solubilized. Therefore, this mechanism is less likely to explain the esophageal mucosal disruption caused by bile acids. Alternatively, the second and more favored hypothesis suggests that bile acids gain entrance across the mucosa because of their lipophilic state, causing intramucosal damage primarily by disorganizing membrane structure or interfering with cellular function. Support for this model comes from several experimental studies. Batzri et al.48 found that bile acids, once penetrating the mucosal barrier, are trapped inside the ceils by intracellular ionization, explaining the severalfold increase in intracellular concentrations of bile acids. 49 Furthermore, studies by Schweitzer e t al. 46'47 have correlated bile acid entry and mucosal accumulation with bile acid-mediated mucosal damage. These findings explain the previous observations of increased mucosal injury by conjugated bile acids at pH 2 and unconjugated bile acids at pH 7. The unionized forms predominate at more acidic pH for conjugated bile acids (pKa 1.9) and at more neutral pH for unconjugated bile acids (pKa 5.1). The unionized forms of the bile acids are more lipophilic, allowing access through the esophageal mucosal barrier into the intracellular compartment, where they are trapped by ionization and subsequently cause mucosal damage.

Human S t u d i e s Role of Acid and Pepsin Early studies in humans concentrated on the damaging effects of acid and pepsin on the esophageal mucosa. The modern concept of peptic esophagitis was first suggested by Winkelstein in 1935, 7° who proposed a

1900

VAEZI ET AL.

role for gastric juice in the formation of esophagitis based on clinical findings in 3 patients. However, studies by Aylwin 51 were the first scientific evidence identifying the importance of acid and pepsin in the development of heartburn and esophageal mucosal injury. Using continuous esophageal aspiration in patients with hiatal hernia and esophagitis, they found that patients with esophagitis had aspirates of lower pH and higher pepsin concentration than those without esophagitis. Later, Turtle et al. 52'53 measured the pH of the distal esophagus, finding that reflux of pH < 4 material coincided with the onset of heartburn, whereas an increase to a more neutral pH coincided with relief of symptoms. Subsequently, a series of studies showed that patients with various grades of esophagitis, including Barrett's esophagus, have increased frequency and duration of esophageal exposure to pH < 4 refluxate. 54-5. Iascone et al. 59 reported a direct relationship between the severity of esophageal mucosal injury and the degree and frequency of mucosal exposure to acid reflux. Later, studies by DeMeester et al. 2 found that 90% of patients with esophagitis had increased amounts of acid reflux by 24hour pH monitoring. The same group 6° reported that patients with Barrett's esophagus had significantly higher exposure times to pH < 4 than patients with esophagitis, who had higher exposure times than healthy controls. Later, Stein et al. 61 reported that patients with Barrett's esophagus had greater exposure times to more caustic gastric acid concentrations (pH < 3 or 2), suggesting a significant role for acid reflux in the development of esophagitis and Barrett's esophagus. Additionally, studies have shown a positive correlation between the degree of abnormal acid and pepsin exposure and the severity of esophagitis. Bremner et al. 62 observed that patients with increased esophageal exposure to pH 0 - 2 , corresponding to the known pKa of pepsin, had the most significant degrees of esophagitis. In a recent study, Gotley et al. 12 found that esophageal aspirates from patients with esophagitis has significantly higher concentrations of acid and pepsin than the aspirates from healthy controls. Furthermore, patients with ZollingerEllison syndrome, in which the basal acid output is high and gastric pH favors optimum acidity for pepsin activity, have a 4 0 % - 6 0 % incidence of esophagitis despite normal or increased lower esophageal sphincter pressures) 3 Contrary to these reports is the data by Hirschowitz, 64 who found no difference between pepsin concentration and acid secretion among patients with esophagitis and healthy subjects. It is important to note that the frequency and duration of esophageal acid exposure is not always predictive of the degree of esophageal mucosal injury. This suggests

GASTROENTEROLOGY Vo1.108, No. 6

the importance of other factors, including the inherent resistance of esophageal mucosa to acid injury and the role of saliva- and bicarbonate-producing submucosal glands in the distal esophagus to neutralize refluxed acid. 65-68

Role of Duodenal Contents Isolated case reports during the last 40 years suggest that duodenal contents alone, without acid or pepsin, may cause esophageal mucosal damage. 69-vl This was postulated by Helsingen, who observed that patients after gastrectomy and resulting achlorhydria could still develop severe esophagitis. 69 Others v°'vl also questioned the role of acid and pepsin as the sole causes of esophageal mucosal damage by reporting heartburn and esophagitis in patients with achlorhydria with and without perni~ cious anemia. The clinical evidence for the possible damaging effects of DGR on the esophageal mucosa remains controversial. This may be because there is no gold standard for detecting DGR. Various direct and indirect methodologies are used for measuring DGR, including endoscopy, aspiration studies (both gastric and esophageal), scintigraphy, ambulatory pH monitoring, and, most recently, ambulatory bilirubin monitoring. As summarized in Table I, these tests have their strengths and shortcomings; however, reviewing the human studies using these tests can help us better appreciate the role of DGR in causing esophageal symptoms and mucosal injury. Endoscopy. Bile is frequently found in the esophagus and stomach of patients during endoscopy; however, the clinical significance of these observations is unclear. Recently, Nasrallah et al. v2 evaluated 110 patients with bile-stained gastric mucosa at endoscopy by measuring gastric bile acids and scintigraphic quantitation of bile reflux. They found no correlation between the gastric bile acid concentrations, degree of histological injury, or severity of endoscopic changes, suggesting that there was little clinical importance to bile-stained mucosa at endoscopy. Similarly, using scintigraphy and gastric pH monitoring to assess DGR, Stein et al. v3 found poor sensitivity (37%), specificity (70%), and positive predictive value (5 5%) for endoscopy in the diagnosis of excessive DGR. Aspiration studies. Stomach. One of the earliest methods used for evaluating DGR into the esophagus was the aspiration of gastric contents with fluid analysis for bile acids. Using this technique, Kaye and Showalterv4 found no significant difference between fasting gastric bile acid concentrations of patients with esophagitis (0.057 mg/mL) compared with controls (0.039 mg/mL), whereas, postprandially, patients with esophagitis had

June 1995

DUODENOGASTRIC REFLUX 1901

Table 1. Advantages and Disadvantages of the Currently Available Methods for Detecting DGR

Method

Advantages

Endoscopy

Easy visualization of bile

Aspiration studies

Less invasive than endoscopy No sedation Low cost Noninvasive

Scintigraphy

pH monitoring

Bilirubin monitoring (Bilitec)

Disadvantages

Easy to perform Relatively noninvasive Prolonged monitoring Ambulatory Easy to perform Relatively noninvasive Prolonged monitoring Ambulatory Good correlation with gastric bile acid concentrations

higher gastric bile acid concentrations (0.89 vs. 0.21 mg/ mL). Similarly, Gillen et al. v5 found no difference in the fasting bile acid concentrations of patients with complicated (strictures, ulcers, dysplasia, adenocarcinoma) (0.025 mmol/L) or uncomplicated (0 mmol/L) Barrett's esophagus compared with patients with esophagitis (0 mmol/L) or normal controls (0 mmol/L). However, they reported significantly higher postprandial bile acid concentrations in patients with complicated Barrett's esophagus (0.19 mmol/L) compared with the other groups (0.017-0.040 mmol/L). The studies finding increased postprandial bile acid concentrations have been criticized because they all used the 3-o~-hydroxysteroid dehydrogenase enzymatic assay, which has recently been reported to have low specificity and accuracy in detecting bile acids in the postprandial but not the fasting state, v6 Recent studies by Vaezi and Richter 7v of patients with complicated and uncomplicated Barrett's esophagus found that fasting bile acid concentrations were higher in complicated (0.5 mmol/L) compared with uncomplicated (0.24 mmol/L) patients with Barrett's esophagus with both concentrations being higher than controls (0.02 mmol/L). These investigators also found that the increased fasting bile acid concentrations in patients with Barrett's esophagus was accompanied by greater amounts of acid and bile reflux, suggesting that both components may be synergistically involved in producing esophageal mucosal damage. A limitation of gastric aspiration studies is the presumption that the presence of bile acids in the stomach is a good indicator of esophageal exposure to duodenal contents and therefore DGR. However, only one half of DGR episodes into the antrum reach the fundus of the

Poor sensitivity/specificity/positive predictive value Requires sedation High cost Short duration of study Requires familiarity with enzymatic assay for bile acid Semiquantitative at best Radiation exposure High cost pH > 7 not a marker for DGR Not specific for DGR

Current design underestimates DGR by about 30% in acidic medium (pH < 3.5) Requires modified diet

stomach, and then all that is present in the fundus may not reflux into the esophagus. 68 Esophagus. Continuous and intermittent esophageal aspiration studies have assessed the role of bile acids in esophageal mucosal damage. Unlike studies of gastric aspirates, most studies with esophageal aspiration technique show very little evidence for the involvement of bile acids in patients with esophagitis. Two independent investigators, Smith et al. v8 and Johnsson et al., v9 found only small concentrations of bile acids (2.5-64 btmol/L) in the esophageal aspirate of patients with GERD. Mittal et al. 76 found no bile acids in either the fasting or postprandial esophageal aspirates of patients with GERD. On the other hand, Gotley et al., 11 studying 45 patients with esophagitis and 10 controls using continuous collection of esophageal aspiration during a period of 16 hours, found increased amounts of conjugated bile acids, measured by high-performance liquid chromatography, in the majority (87%) of aspirates. Most bile acid reflux occurred at night with 7% of samples having bile acid concentrations above 1.0 mmol/L, the toxic concentration producing esophageal mucosal damage. Interestingly, acid reflux episodes measured by 24-hour pH monitoring occurred concomitantly with reflux of conjugated bile acids. However, a second study by the same group 12 later found that the esophageal aspirates of patients with esophagitis only rarely (2%) had conjugated bile acid concentrations high enough (>1.0 mmol/L) to cause esophageal mucosal damage. Additionally, they found no unconjugated bile acids or trypsin in the aspirates, whereas acid and pepsin was found in almost all specimens. Thus, supporting the animal studies, they concluded that reflux esophagitis was caused by acid and

1902

VAEZI ET AL.

pepsin with bile acids and trypsin having insignificant roles. Scintigraphy. Radionuclide techniques offer a noninvasive method for studying DGR. Tolin et al. 8° compared D G R measured by phenol red aspiration and external scintigraphy measuring 99mTc-hepato-iminodiacetic acid scan, finding a good correlation between DGR and scintigraphy. Similarly, other investigators s1'82 have compared intragastric bile acid concentrations and scintigraphy, finding a significant correlation between both free and total gastric bile acids and the degree of bile reflux. Scintigraphic studies find that DGR is a common phenomenon in normal individuals postprandially, 7'a requiring that the evaluation of abnormal DGR be quantitative. Matikainen et al. 83 found no difference in the scintigraphic amount of DGR into the esophagus between 40 patients with esophagitis (10% scintigraphic reflux) and 150 healthy controls (14% scintigraphic reflux). Likewise, Krog et al. 84 found no evidence of DGR in 15 patients with hiatal hernia and esophagitis. However, Warring et al. 85 report that patients with Barrett's esophagus, especially those with complicated Barrett's, had more frequent DGR detected by 99mTc-diisopropyl iminodiacetic acid scintigraphy than healthy volunteers. Although less invasive than other methods for detecting DGR, the reliability and accuracy of scintigraphy has been challenged. Drane et al. .6 observed that scintigraphy is at best a semiquantitative measure of bile reflux, finding that several technical problems may compromise the accuracy of this technique for measuring DGR. The most common problem was the overlap of small bowel and stomach occurring in 36% of patients, which is not correctable. Other problems included overlap of the left lobe of the liver and stomach, patient movement, and the intermittent nature of bile reflux. Ambulatory pH monitoring. Prolonged pH monitoring offers a unique opportunity for studying acid and possibly DGR in the ambulatory state throughout the circadian cycle. Stein et al. 87 reported that gastric monitoring o f p H > 7 is superior to diisopropyl iminodiacetic acid radionucleotide scintigraphy in detecting DGR in patient with foregut symptoms. Similarly, Brown et al. 88 observed a good correlation (r = 0.36; P < 0.001) between gastric bile acid concentrations and ambulatory gastric pH monitoring. Using 24-hour esophageal pH monitoring, Pellegrini e t al. 89 were the first to study the relationship between acid and alkaline reflux in patients with GERD. Acid reflux was defined as pH < 4 in the lower esophagus, whereas alkaline reflux, an indirect marker of DGR, was defined as pH > 7. Normal values were defined as the

GASTROENTEROLOGY Vol. 108, No. 6

mean and 2 SD of values obtained during 24-hour pH studies in 15 healthy volunteers. Compared with patients with acid reflux, alkaline refluxers had less heartburn but more frequent and severe regurgitation. Additionally, there were no pure alkaline refluxers because all patients also had episodes of acid reflux. Later, the same group 9°'91 found that patients with esophagitis had less alkaline reflux and more acid reflux than patients without esophagitis. In contrast, Schmid e t al., 92 studying patients with various degrees of esophagitis, reported significantly higher amounts of both acid and alkaline reflux in patients with complicated esophagitis compared with healthy subjects. Their results, similar to the findings in gastric aspiration studies, suggest a synergistic role for acid and bile in esophagitis. At the same time, Attwood et al. 93-96 reported that alkaline reflux was greater in patients with Barrett's esophagus when compared with patients with esophagitis or normal controls. Furthermore, they found that pH > 7 was significantly higher in patients with complicated Barrett's esophagus (stricture, ulcer, dysplasia) than patients with Barrett's esophagus without complications, whereas pH < 4 did not distinguish the two groups. These investigators went on to suggest that prolonged exposure to duodenal contents alone may promote the development of complicated Barrett's esophagus and even adenocarcinoma. However, the investigators did not attempt to resolve the physiological paradox of patients with marked amount of acid reflux also refluxing, independently, large amounts of alkaline material. The measurement of esophageal pH > 7 as a marker of DGR is confounded by several problems. Precautions must be taken to use only glass electrodes, dietary restriction of foods with pH < 7, inspection of patients for periodontal disease, and dilation of strictures to avoid pooling of saliva. Therefore, it is not surprising that several investigators 66-68'97 have questioned the accuracy of alkaline pH as a parameter for monitoring duodenal reflux into the esophagus. Gotley et al. 97 found no relationship between alkaline exposure time and esophageal bile acids or trypsin. Similarly, Mattioli et al., 68 using a triple-probe pH monitor placed in the distal esophagus, fundus, and antrum, found that alkaline reflux, defined as an increase in pH > 7 from the antrum to esophagus, was extremely uncommon (0.75%) in 279 patients. Therefore, the investigators suggested that increases of esophageal pH above 7 were most likely due to other reasons (saliva, food, oral infection, obstructed esophagus) rather than reflux of duodenal contents. This speculation was substantiated by studies from Singh et al. 66 and DeVault et al.,67 who found that increased saliva production or bicarbonate production by the esophageal submu-

June 1995

cosal glands were the most common cause of esophageal pH > 7, whereas DGR was rare in patients with intact stomach. Finally, using an ambulatory bilirubin monitoring device, Champion et al. 98 reported no correlation between alkaline reflux and bile reflux into the esophageal lumen, suggesting that the term "alkaline" reflux was a misnomer and should not be used when referring to DGR. Ambulatory bilirubin monitoring (Bilitec 2000). Recently, a new fiber-optic spectrophotometer (Bilitec 2000; Synectics, Stockholm, Sweden) was developed that detects DGR in an ambulatory setting, independent of pH. 99 The system consists of a miniaturized fiber-optic probe that carries light signals into the esophagus and back to the optoelectronic system via a plastic fiber-optic bundle. Two light emitting diodes (at 470 and 565 nm) represent the sources for the measurement of bilirubin, the most common pigment in bile, and the reference signal (i.e., no bilirubin absorption), respectively. The difference in absorption between the two wavelengths is proportional to bilirubin concentration in the sample under study. Several reports have indicated a good correlation between Bilitec readings and bile acid concentration measured by duodenogastric aspiration studies (r = 0.71, P < 0.0196 and r = 0.82, P < 0.001). 1°° This spectrophotometric technique may be an important advancement in the assessment of DGR, permitting more accurate studies of patients with syndromes associated with DGR. Additionally, it may be used concomitantly with pH monitoring to measure the esophageal exposure to both acid and DGR because some of the previously discussed methodologies suggest that concomitant reflux of both acid and bile may be more common than previously thought. There are only a few studies using the Bilitec system to evaluate patients with esophageal mucosal injury. In a preliminary report, Kauer et al. I°1 found no significant difference in esophageal bilirubin exposure in patients with esophagitis compared with healthy controls; however, patients with Barrett's esophagus had significantly more bilirubin and acid reflux than controls. On the other hand, Champion et al. 98 found a significant but graded increase in both acid and DGR from controls to patients with esophagitis with the highest values observed in patients with Barrett's esophagus. Furthermore, DGR had a strong correlation with acid reflux (r = 0.78) but had a poor association with pH > 7 (r = -0.06). Further support for the graduated increase in both acid and DGR comes from recent studies by Vaezi and Richter vv of patients with and without complications of Barrett's esophagus. They found that both groups of patients with Barrett's esophagus refluxed significantly greater

DUODENOGASTRIC REFLUX 1903

quantities of bile and acid into their lower esophagus than controls. More importantly, reflux of acid paralleled DGR, and both were significantly higher in patients with complicated Barrett's than the uncomplicated group. Thus, the preliminary data from the Bilitec studies suggest a possible synergy between bile and acid in the development of esophagitis and Barrett's esophagus. However, because of inherent limitations of the Bilitec in quantitating DGR, future studies are needed to better define the degree of synergy between DGR and acid reflux in esophageal mucosal injury. Bilitec 2000 is an important new tool in detecting DGR; however, as currently designed, it has limitations. Validation studies by Vaezi et al. 1°° found that this instrument underestimates bile reflux by at least 30% in an acidic medium (pH < 3.5). In solutions with pH < 3.5, bilirubin undergoes monomer to dimer isomerization, which is reflected by the shift in the absorption wavelength from 453 to 400 nm. Because Bilitec readings are based on the detection of absorption at 470 nm, this shift results in underestimation of the degree of DGR. Therefore, further design modifications are necessary to optimize the quantitative detection of DGR by this instrumentation. The design modifications could include changing the diode wavelength to 390 nm, representing the wavelength that is unaffected by the bilirubin monomer to dimer transition, or using a dual sample detection diode, one at 450 nm for refluxate with pH > 3.5 and another diode at 400 nm for pH < 3.5 refluxate. Furthermore, a variety of substances may result in falsepositive readings by the Bilitec because it indiscriminately records any substance absorbing around 470 nm. This necessitates use of a modified diet to avoid interference and false readings. 99'I°° Also, it is important to remember that Bilitec measures reflux of bilirubin and not bile acids, thereby presuming that the presence of bilirubin in the refluxate is accompanied by other duodenal contents. Although this is true in most cases, a few medical conditions (Gilbert's and Dubin-Johnson syndromes) may result in disproportionate secretion of bilirubin as compared with other duodenal contents, especially bile acids. 102 Clinical Implications Both animal and human studies suggest that acid is the key agent in causing esophageal mucosal injury. However, studies now suggest that duodenogastric contents (bile acids, trypsin, lysolecithin) also frequently reflux into the esophagus along with stomach acid. Animal studies suggest that DGR, at concentrations of > 1 mmol/L, usually causes the most esophageal damage in synergy with acid. 25'32 This was recently confirmed in

1904

VAEZI ET AL.

human studies by Sears et al. 1°3 in patients with reflux symptoms after duodenal ulcer surgery. This patient population represents an excellent human model for increased DGR because of the incompetent pylorus and free regurgitation of duodenal contents into the stomach, resulting in gastric bile acids concentrations (0.5-3.0 mmol/L) known to cause esophageal mucosal injury in the animal model (>1 mmol/L). These investigators studied 13 patients with partial gastrectomy with reflux symptoms, finding increased DGR by Bilitec monitoring in 77% of patients. However, endoscopic esophagitis was only present in those with concomitant acid reflux. Therefore, supporting the animal reports, this study underscores the important point that DGR, without abnormal amounts of acid reflux, may cause reflux symptoms but does not usually result in esophageal mucosal injury. It is well known that aggressive medical therapy to suppress acid secretion will heal most cases ofesophagitis. Recently, some groups 56'93-96 have suggested that prolonged acid suppression in patients with severe esophagitis or Barrett's esophagus may promote DGR, causing further mucosal injury with progression to Barrett's esophagus or even adenocarcinoma. This claim is based on isolated animal studies *°4'1°5 and a handful of clinical reports, mainly from the same laboratory. 55'93-95 The clinical relevance of these reports is questionable because the association between DGR and complicated Barrett's esophagus or adenocarcinoma was defined by esophageal pH monitoring (i.e., pH > 7), which has since been shown to be an unreliable parameter for detection of DGR. More importantly, a careful review of the literature shows that the overwhelming majority of studies do not support this concept. Both animal and human studies indicate that DGR in the absence of acid reflux is usually not damaging to the esophageal mucosa. Furthermore, recent studies by Champion et al. 98 in 9 patients with severe GERD (esophagitis, 3; Barrett's esophagus, 6) found that aggressive acid suppression with omeprazole (20 mg twice daily) dramatically decreased both acid and DGR (Figure 2). Although not specifically studied, the investigators speculate this was due to omeprazole's inhibition of both gastric acidity and volume. Thus, despite a defective lower esophageal sphincter, less gastric volume was available to reflux into the esophagus throughout the circadian cycle, regardless of the duodenogastric contents. Therefore, medical therapy with aggressive acid suppression may not only protect the esophageal mucosa from the damaging effects of acid and eliminate the synergy between acid and bile but also decrease the volume of both acid and bile refluxing into the esophagus. The findings by Champion et al. 9. have important implications for treating patients with both acid and bile

GASTROENTEROLOGY Vol. 108, No. 6

25

A

20 V "1O.

15

E I-m

P-

10

5

0 Pretreatment

Omeprazole

60 c5 AI (D 0 t-

40 O o3

<

x

c-

l

. m

.Q . m

rn o

20

E

. m

#0 Pretreatment

Omeprazole

Figure 2. Influence of marked acid suppression with 20 mg omeprazole twice daily on (A) acid and (B) duodenogastroesophageal reflux in 9 patients with severe gastroesophageal reflux disease. Mean values for both parameters significantly (P < 0.001) decreased from baseline values by the end of 2 weeks of therapy. These changes were found for all patients regardless of their pretreatment values (Reprinted with permission.98),

June 1995

reflux. Previously the best therapy for decreasing DGR into the esophagus was thought to be an antireflux operation correcting the defective lower esophageal sphincter. However, this study suggests that medical therapy may decrease both acid and DGR to a similar degree as antireflux surgery. Medical therapy has the advantage of avoiding a surgical procedure and its associated complications, an important consideration in the elderly and those with contraindications to surgery. However, in younger patients in whom long-term medical therapy is anticipated, antireflux surgery may be a more suitable and cost-effective alternative.

References 1. The Gallop Organization: A Gallop survey on "heartburn across America." Princeton, NJ: The Gallup Organization, 1988. 2. DeMeester TR, Wernly JA, Little AG, Bermudez G, Skinner DB. Technique, indications, and clinical use of 24 hour esophageal pH monitoring. J Thorac Cardiovasc Surg 1980; 7 9 : 6 5 6 - 6 7 0 . 3. Redo SF, Barnes WA, de la Sierra AO. Perfusion of the canine esophagus with secretions of the upper gastro-intestinal tract. Ann Surg 1 9 5 9 ; 1 4 9 : 5 5 6 - 5 6 4 . 4. Ferguson DJ, Sanchez-Palomera E, Sako Y, Caltworthy HW, Toon RW, Wangensteen OH. Studies on experimental esophagitis. Surgery 1950; 2 8 : 1 0 2 2 - 1 0 3 9 . 5. Kiriluk LB, Merendino KA. Comparative sensitivity of mucosa of various segments of alimentary tract in dog to acid-peptic action. Surgery 1954; 3 5 : 5 4 7 - 5 5 6 . 6. Goldberg HI, Dodds WJ, Gee S, Montgomery C, Zboralske FF. Role of acid and pepsin in acute experimental esophagitis. Gastroenterology 1969; 5 6 : 2 2 3 - 2 3 0 . 7. Mackie C, Hulks G, Cuschieri A. Enterogastric reflux and gastric clearance of refluxate in patients with and without bile vomiting following peptic ulcer surgery. Ann Surg 1 9 8 6 ; 2 0 4 : 5 3 7 - 5 4 2 . 8. Muller-Lissner SA, Fimmel CJ, Sonnenberg A. Novel approach to quantify duodenogastric reflux in healthy volunteers and in patients with type I gastric ulcer. Gut 1 9 8 3 ; 2 4 : 5 1 0 - 5 1 8 . 9. Schidlbeck NE, Heinrich C, Stellard F, Paumgartner G, MullerLissner SA. Healthy controls have as much bile reflux as gastric ulcer patients. Gut 1987; 2 8 : 1 5 7 7 - 1 5 8 3 . 10. King PM, Pryde A, Heading RC. Transpyloric fluid movement and antroduodenal motility in patients with gastroesophageal reflux. Gut 1987; 2 8 : 5 4 5 - 5 4 8 . 11. Gotley DC, Morgan AP, Cooper MJ. Bile acid concentrations in the refluxate of patients with reflux oesophagitis. Br J Surg 1988;75:587-590. 12. Gotley DC, Morgan AP, Ball D, Owens RW, Cooper M J. Composition of gastro-oesophageal refluxate. Gut 1 9 9 1 ; 3 2 : 1 0 9 3 1099. 13. Burden WR, Hodges RP, Hsu M, O'Leary JP. Alkaline reflux gastritis. Surg Clin North Am 1991; 7 1 : 3 3 - 4 4 . 14. Dixon MF. Progress in pathology of gastritis and duodenitis. In: Williams GT, ed. Current topics in pathology. Berlin: SpringerVerlag; 1 9 9 0 : 1 - 3 9 . 15. DuPlessis DJ. Pathogenesis of gastric ulceration. Lancet 1965; 1 : 9 7 4 - 9 7 8 . 16. Ritchie WP. Alkaline reflux gastritis: a critical reappraisal. Gut 1984;25:975-987. 17. Harmon JW, Bass BL, Batzri S. Alkaline reflux gastritis. In: Nyhus LL, ed. Problems in general surgery. Philadelphia: Lippincott, 1993:201-206. 18. Hopewood D, Bateson MC, Milner G, Bouchier lAD. Effects of

DUODENOGASTRIC REFLUX

19.

20. 21.

22.

23. 24.

25. 26. 27.

28.

29. 30.

31.

32.

33.

34.

35.

36. 37. 38.

39.

40.

41.

1905

bile acids and hydrogen ion on the fine structure of oesophageal epithelium. Gut 1981; 2 2 : 3 0 6 - 3 1 1 . Hubens A, Van de Kelft E, Roland J. The influence of cholecystectomy on the duodenogastric reflux of bile. Hepatogastroenterol 1989; 3 6 : 3 8 4 - 3 8 6 . Kalima TV. Reflux gastritis unrelated to gastric surgery. Scand J Gastroenterol 1982; 17(suppl 7 9 ) : 6 6 - 7 1 . Muller-Lissner SA, Schindlbeck NE, Heinrich C. Bile salt reflux after cholecystectomy. Scand J Gastroenterol 1987;22(suppl 139):20-24. Nano M, Palmas F, Giaccone M, Ricci E, Camandona M, Gardino L, Coluccia C, Aloesio R, Fausone G. Biliary reflux after cholecystectomy: A prospective study. Hepatogastroenterol 1990;37: 233-234. Stoker DL, Williams JG. Alkaline reflux oesophagitis. Gut 1991; 3 2 : 1 0 9 0 - 1 0 9 2 . Lillemoe KD, Johnson LE, Harmon JW. Alkaline esophagitis: a comparison of the ability of components of gastroduodenal contents to injure rabbit mucosa. Gastroenterology 1983; 85:621-628. Salo J, Kivilaaksc E. Role of luminal H ÷ in the pathogenesis of experimental esophagitis. Surgery 1 9 8 2 ; 9 2 : 6 1 - 6 8 . Harmon JW, Doong T, Gadacz TR. Bile acids are not equally damaging to the gastric mucosa. Surgery 1 9 7 8 ; 8 4 : 7 9 - 8 6 . Lillemoe KD, Johnson LF, Harmon JW. Role of the components of the gastroduodenal contents in experimental acid esophagitis. Surgery 1 9 8 2 ; 9 2 : 2 7 6 - 2 8 4 . Cross FS, Wangensteen OH. Role of bile and pancreatic juice in the production of esophageal erosions and anemia. Proc Soc Exp Biol Med 1961; 7 7 : 8 6 2 - 8 6 6 . Moffat RC, Berkas EM. Bile esophagitis. Arch Surg 1965; 91:963-966. Harmon JW, Johnson LF, Maydonovitch CL. Effects of acid and bile salts on the rabbit esophageal mucosa. Dig Dis Sci 1981;26:65-72. Kivilaakso E, Fromm D, Silen W. Effect of bile salts and related compounds on isolated esophageal mucosa. Surgery 1980; 87:280-285. Gillison EW, DeCastro VAM, Nyhus LM, Kusakari K, Bombeck CT. The significance of bile in reflux esophagitis. Surg Gynecol Obstet 1972; 1 3 4 : 4 1 9 - 4 2 4 . Salo JA, Kivilaakso E. Contribution of trypsin and cholate to the pathogenesis of experimental alkaline reflux esophagitis. Scand J Gastroenterol 1984; 1 9 : 8 7 5 - 8 8 1 . Lillemoe KD, Johnson LF, Harmon JW. Taurodeoxycholate modulates the effects of pepsin and trypsin in experimental esophagitis. Surgery 1 9 8 5 ; 9 7 : 6 6 2 - 6 6 7 . Kiroff GK, Mukerjhee TM, Dixon B, Devitt PG, Jamieson GG. Morphological changes caused by exposure of rabbit oesophageal mucosa to hydrochloric acid and sodium taurocholate. Aust N Z J Surg 1 9 8 7 ; 5 7 : 1 1 9 - 1 2 6 . Salo JA, Lehto VP, Kivilaakso E. Morphological alteration in experimental esophagitis. Dig Dis Sci 1 9 8 3 ; 2 8 : 4 4 0 - 4 4 8 . Shimono M, Clementi F. Intercellular junction of oral epithelium. J Ultrastruct Res 1 9 7 7 ; 5 9 : 1 0 1 - 1 1 2 . Carney CN, Orlando RC, Powell DW. Morphologic alterations in early acid induced epithelial injury of the rabbit esophagus. Lab Invest 1 9 8 1 ; 4 5 : 1 9 8 - 2 0 8 . Orlando RC, Bryson JC, Powell DW. Mechanism of H + injury in rabbit esophageal epithelium. Am J Physiol 1984;9:G718G724. Orlando RC, Powell DW, Carney CN. Pathophysiology of acute acid injury in rabbit esophageal epithelium. J Clin Invest 1981;68:286-293. Powell DW, Orlando RC, Carney CN. Acid injury of the esophageal epithelium. In: Harmon JW, ed. Basic mechanism of gastro-

1906

42.

43.

44.

45.

46.

47.

48.

49.

50. 51. 52.

53. 54.

55.

56. 57. 58.

59.

60. 61.

62.

63.

64.

VAEZI ET AL.

intestinal mucosal cell injury and protection. Baltimore, MD: Williams & Wilkins, 1981:55. Snow JC, Goidstein JL, Schmidt LN, Lisitza P, Layden TJ. Rabbit esophageal cells show regulatory volume decrease: ionic basis and effect of pH. Gastroenterology 1 9 9 3 ; 1 0 5 : 1 0 2 - 1 1 0 . Duane WC, Wiegand DM. Mechanism bywhich bile salt disrupts the gastric mucosal barrier in the dog. J Clin Invest 1980; 6 6 : 1 0 4 4 - 1 0 4 9 . Thomas AJ, Nahrwold DL, Rose RC. Detergent action of sodium taurocholate on rat gastric mucosa. Biochim Biophys Acta 1972; 2 8 2 : 2 1 0 - 2 1 3 . Tanaka K, Fromm F. Effect of bile acid and salicylate on isolated surface and glandular cells of rabbit stomach. Surgery 1983; 9 3 : 6 6 0 - 6 6 3 . Schweitzer EJ, Harmon JW, Bass BL, Batzri S. Bile acid efflux precedes mucosal barrier disruption in the rabbit esophagus. Am J Physiol 1984;10:G480-G485. Schweitzer EJ, Bass BL, Batzri S, Young PM, Huesken J, Harmon JW. Lipid solubilization during bile salt-induced esophageal mucosal barrier disruption in the rabbit. J Lab Clin Med 1987; 1 1 0 : 1 7 2 - 1 7 9 . Batzri S, Harmon JW, Schweitzer EJ, Toles R. Bile acid accumulation in gastric mucosal cells. Proc Soc Exp Biol Med 1991;197:393-399. Schweitzer EJ, Bass BL, Batzri S, Harmon JW. Bile acid accumulation by rabbit esophageal mucosa. Dig Dis Sci 1986; 31:1105-1113. Winkelstein A. Peptic esophagitis. A new clinical entity. JAMA 1935; 1 0 4 : 9 0 6 - 9 0 9 . Aylwin JA. The physiological basis of reflux esophagitis in sliding distal diaphragmatic hernia. Thorax 1 9 5 3 ; 8 : 3 8 - 4 5 . Turtle SG, Bettarello A, Gossman MI. Esophageal acid perfusion test and a gastroesophageal reflux in patients with esophagitis. Gastroenterology 1 9 6 0 ; 3 8 : 8 6 1 - 8 7 2 . Tuttle SG, Rufin F, Bettarelio A. The physiology of heartburn. Ann Intern Med 1 9 6 1 ; 5 5 : 2 9 2 - 3 0 0 . Zamost BJ, Hirschberg J, Ippoliti AF. Esophagitis in scleroderma: prevalence and risk factors. Gastroenterology 1987; 92:421-428. Stein HJ, Barlow AP, DeMeester TR, Hinder RA. Complications of gastroesophageal reflux disease. Ann Surg 1 9 9 2 ; 2 1 6 : 3 5 43. Hennessy TPJ. Barrett's esophagus. Br J Surg 1985; 7 2 : 3 3 6 340. Giiien P, Keeling P, Byrne PJ, Hennessy TPJ. Barrett's esophagus: pH profile. Br J Surg 1 9 8 7 ; 7 4 : 7 7 4 - 7 7 6 . Stein HJ, Siewert JR. Barrett's esophagus: pathogenesis, epidemiology, functional abnormalities, malignant degeneration, and surgical management. Dysphagia 1 9 9 3 ; 8 : 2 7 6 - 2 8 8 . lascone C, DeMeester TR, Little AG, Skinner DB. Barrett's esophagus: functional assessment, proposed pathogenesis, and surgical therapy. Arch Surg 1983; 118:543-549. Stein HJ, Hoeft S, DeMeester TR. Reflux and motility pattern in Barrett's esophagus. Dis Esoph 1 9 9 2 ; 5 : 2 1 - 2 8 . Stein HJ, Hoeft S, DeMeester TR. Functional foregut abnormalities in Barrett's esophagus. J Thorac Cardiovasc Surg 1993; 1 0 5 : 1 0 7 - 1 1 1 . Bremner RM, Crookes PF, DeMeester TR, Peters JH, Stein H. Concentration of refluxed acid and esophageal mucosal injury. Am J Surg 1992; 1 6 4 : 5 2 2 - 5 2 7 . Miller LS, Vinayek R, Frucht H, Gardner JD, Jensen RT, Maton PN. Reflux esophagitis in patients with Zollinger-Ellison syndrome. Gastroenterology 1990; 9 8 : 3 4 1 - 3 4 5 . Hirschowitz BI. A critical analysis, with appropriate controls, of gastric acid and pepsin secretion in clinical esophagitis. Gastroenterology 1991; 1 0 1 : 1 1 4 9 - 1 1 5 8 .

GASTROENTEROLOGY Vol. 108, No. 6

65. Hamilton BH, Orlado RC. In vivo alkaline secretion by mammalian esophagus. Gastroenterology 1 9 8 9 ; 9 7 : 6 4 0 - 6 4 8 . 66. Singh S, Bradley LA, Richter JE. Determinants of oesophageal "alkaline" pH environment in controls and patients with gastrooesophageal reflux disease. Gut 1 9 9 3 ; 3 4 : 3 0 9 - 3 1 6 . 67. Devault KR, Georgeson S, Castell DO. Salivary stimulation mimics esophageal exposure to refluxed duodenal contents. Am J Gastroenterol 1 9 9 3 ; 8 8 : 1 0 4 0 - 1 0 4 3 . 68. Mattioli S, Pilotti V, Felice V, Lazzzari A, Zannoli R, Bachi ML. Ambulatory 24 hour pH monitoring of the esophagus, fundus and antrum. Dig Dis Sci 1 9 9 0 ; 3 5 : 9 2 9 - 9 3 8 . 69. Helsingen N. Oesophagitis following total gastrectomy. A follow up study on 9 patients 5 years or more after operation. Acta Chir Scand 1959; 1 1 8 : 1 9 0 - 2 0 1 . 70. Orlando RC, Bozymski EM. Heartburn in pernicious a n e m i a - a consequence of bile reflux. N Engl J Med 1 9 7 3 ; 2 8 9 : 5 2 2 523. 71. Palmer ED. Subacute erosive ("peptic") esophagitis associated with achlorhydria. N Engl J Med 1 9 6 0 ; 2 6 2 : 9 2 7 - 9 2 9 . 72. Nasrallah SM, Johnston GS, Gadacz TR, Kim KM. The significance of gastric bile reflux seen at endoscopy. J Clin Gastroenterol 1 9 8 7 ; 9 : 5 1 4 - 5 1 7 . 73. Stein HJ, Smyrk TC, DeMeester TR, Rouse J, Hinder RA. Clinical value of endoscopy and histology in the diagnosis of duodenogastric reflux disease. Surgery 1992; 1 1 2 : 7 9 6 - 8 0 4 . 74. Kaye MD, Showalter JP. Pyloric incompetence in patients with symptomatic gastroesophageal reflux. J Lab Clin Med 1974; 8 3 : 1 9 8 - 2 0 6 . 75. Gillen P, Keeling P, Byrne PJ, Healy M, O'Moore RR, Hennessy TPJ. Importance of duodenogastric reflux in the pathogenesis of Barrett's oesophagus. Br J Surg 1 9 8 8 ; 7 5 : 5 4 0 - 5 4 3 . 76. Mittal RK, Reuben A, Whitney JO, McCallum RW. Do bile acids reflux into the esophagus? A study in normal subject and patients with GERD. Gastroenterology 1987; 9 2 : 3 7 1 - 3 7 5 . 77. Vaezi MF, Richter JE. Complicated Barrett's esophagus: role of acid and bile (abstr). Am J Gastroenterol 1 9 9 4 ; 8 9 : 1 6 3 0 (A62). 78. Smith MR, Buckton GK, Bennett JR. Bile acid levels in stomach and oesophagus of patients with acid gastroesophageal reflux (abstr). Gut 1984;25:A556. 79. Johnsson F, Joelsson B, Floren CH, Nilsson A. Bile salts in the esophagus of patients with esophagitis. Scand J Gastroenterol 1988;23:712-716. 80. Tolin RD, Malmud LS, Stelzer F, Menin R, Makler PT, Applegate G, Fisher RS. Enterogastric reflux in normal subjects and patients with Bilroth II gastroenterostomy. Measurement of enterogastric reflux. Gastroenterology 1979; 7 7 : 1 0 2 7 - 1 0 3 3 . 81. Eriksson B, Emas S, Jacobsson H, Larsson SA, Samuelsson K. Comparison of gastric aspiration and HIDA scintigraphy in detecting fasting duodenogastric bile reflux. Scand J Gastroenterol 1988; 2 3 : 6 0 7 - 6 1 0 . 82. Houghton PWJ, McMortensen NJ, Thomas WEG, Cooper M J, Morgan AP, Davies ER. Intragastric bile acids and scintigraphy in the assessment of duodenogastric reflux. Br J Surg 1986; 7 3 : 2 9 2 - 2 9 4 . 83. Matikainen M, Taavitsainen M, Kalima TV. Duodenogastric reflux in patients with heartburn and esophagitis. Scand J Gastroenterol 1981; 1 6 : 2 5 3 - 2 5 5 . 84. Krog M, Gustavsson S, Jung B. Studies on oesophagitis--No evidence for pyloric incompetence as a primary etiological factor. A scintigraphic study with 99Tcm-Solco-HIDA. Acta Chir Scand 1982; 1 4 8 : 4 3 9 - 4 4 2 . 85. Warring JP, Legrand J, Chinichian A, Sanowski RA. Duodenogastric reflux in patients with Barrett's esophagus. Dig Dis Sci 1990; 35: 7 5 9 - 762. 86. Drane WE, Karvelis K, Johnson DA, Silverman ED. Scintigraphic evaluation of duodenogastric reflux. Problems, pitfalls, and technical review. Clin Nucl Med 1987; 1 2 : 3 7 7 - 3 8 4 .

June 1995

87. Stein HJ, Hinder RA, DeMeester TR, Lloyd BA, Fuchs KH, Attwood EA, Gupta NC. Clinical use of 24-hour gastric monitoring vs o-diisopropyl iminodiacetic acid (DISIDA) scanning in the diagnosis of pathologic duodenogastric reflux. Arch Surg 1990; 125:966-971. 88. Brown TH, Holbrook I, King RFG, Ibrahim K. 24-hour intragastric pH measurement in the assessment of duodenogastric reflux. World J Surg 1992; 16:995-1000. 89. Pellegrini CA, DeMeester TR, Wernly JA, Johnson LF, Skinner DB. Alkaline gastroesophageal reflux. Am J Surg 1978; 75:177184. 90. Little AG, DeMeester TR, Kirchner PT, O'Sullivan GC, Skinner DB. Pathogenesis of esophagitis in patients with gastroesophageal reflux. Surgery 1980;88:101-107. 91. Little AG, Martinez El, DeMeester TR, Blough RM, Skinner DB. Duodenogastric reflux and reflux esophagitis. Surgery 1984; 96:447 -454. 92. Schmid B, DeTarnowsky G, Layden T. Alkaline gastroesophageal reflux--role in esophageal injury (abstr). Gastroenterology 1981; 80:A1275. 93. Attwood SEA, DeMeester TR, Bemner CG, Barlow AP, Hinder RA. Alkaline gastroesophageal reflux: implications in the development of complications in Barrett's columnar-lined lower esophagus. Surgery 1989; 106:764-776. 94. Attwood SEA, Barlow AP, Norris TL, Watson A. Barrett's esophagus: effect of antireflux surgery on symptom control and development of complications. Br J Surg 1992;79:1050-1053. 95. Attwood SEA, Ball CS, Barlow AP, Jenkinson L, Norris TL, Watson A. Role of intragastric and intraoesophageal alkalinization in the genesis of complications in Barrett's columnar lined lower oesophagus. Gut 1993; 34:11-15. 96. Attwood SEA. New advances in the understanding of Barrett's esophagus. Surg Ann 1993; 25:151-175. 97. Gotley DC, Appleton GVN, Cooper MJ. Bile acids and trypsin are

DUODENOGASTRIC REFLUX 1907

98.

99.

100.

101.

102.

103.

104.

105.

unimportant in alkaline esophageal reflux. J Clin Gastroenterol 1992; 14:2-7. Champion G, Richter JE, Vaezi MF, Singh S, Alexander R. Duodenogastroesophageal reflux: relationship to pH and importance in Barrett's esophagus. GastroenterologS' 1994;107:747754. Bechi P, Paucciani F, Baldini F, Cosi F, Falciai R, Mazzanti R. Castagnli A, Pesseri A, Boscherini S. Long-term ambulatory enterogastric reflux monitoring. Validation of a new fiberoptic technique. Dig Dis Sci 1993;38:1297-1306. Vaezi MF, LaCamera RG, Richter JE. Bilitec 2000 ambulatory duodenogastric reflux monitoring system. Studies on its validation and limitations. Am J Physiol 1994;30:1050-1056. KauerWK, Ireland AP, Burdiles P, Clark GWB, Peters JH, DeMeester TR. The correlation between bile and acid reflux in Barrett's esophagus (abstr). Gastroenterology 1994; 106:A104. Lumeng L, O'Connor KW. Differential diagnosis of jaundice. In: Ostrow JD, ed. Bile pigments and jaundice. New York: Marcel Dekker, 1986:475-538. Sears RJ, Champion G, Richter JE. Characteristics of partial gastrectomy (PG) patients with esophageal symptoms of duodenogastric reflux. Am J Gastroenterol 1995;90:211-215. Attwood SEA, Smyrk TC, DeMeester TR, Mirvish SS, Stein HJ, Hinder RA. Duodenoesophageal reflux and the development of esophageal adenocarcinoma in rats. Surgery 1992; 111:503510. Pera M, Trastek VF, Carpenter HA, Fernandez PL, Cardesa A, Mohr Ulrich, Pairolero PC. Influence of pancreatic and biliary reflux on the development of esophageal carcinoma. Ann Thorac Surg 1993; 55:1386-1393.

Received February 7, 1995. Accepted March 14, 1995. Address requests for reprints to: Joel E. Richter, M.D., Department of Gastroenterology, The Cleveland Clinic Foundation/S40, 9500 Euclid Avenue, Cleveland, Ohio 44195. Fax: (216) 444-9416.