ANNALS OF SURGERY REVIEW Vol. 231, No. 3, 303–321 © 2000 Lippincott Williams & Wilkins, Inc.

Columnar Mucosa and Intestinal Metaplasia of the Esophagus Fifty Years of Controversy

Steven R. DeMeester, MD,* and Tom R. DeMeester, MD†

From the Departments of *Cardiothoracic Surgery and †Surgery, The University of Southern California School of Medicine, Los Angeles, California

Objective esophageal reflux. Intestinal metaplasia, even a short To outline current concepts regarding etiology, diagnosis, and length, is premalignant, and the presence of dysplasia indi- treatment of intestinal metaplasia of the esophagus and cardia. cates progression on the pathway to adenocarcinoma. An- tireflux surgery, as opposed to medical therapy, may in- Summary Background Data duce regression or halt progression of intestinal Previously, endoscopic visualization of columnar mucosa ex- metaplasia. The presence of high-grade dysplasia is fre- tending a minimum of 3 cm into the esophagus was sufficient quently associated with an unrecognized focus of adeno- for the diagnosis of Barrett’s esophagus, but subsequently carcinoma. Vagal-sparing esophagectomy removes the the importance of intestinal metaplasia and the premalignant diseased esophagus and is curative in patients with high- nature of Barrett’s have been recognized. It is now apparent grade dysplasia. Invasion beyond the mucosa is associated that shorter lengths of intestinal metaplasia are common, and with a high likelihood of lymph node metastases and re- share many features of traditional 3-cm Barrett’s esophagus. quires lymphadenectomy. Methods Themes and concepts pertaining to intestinal metaplasia of the esophagus and cardia are developed based on a review Conclusions of the literature published between 1950 and 1999. Despite improved understanding of this disease, controversy about the definition and best treatment of Barrett’s esopha- Results gus continues, but new molecular insights, coupled with care- Cardiac mucosa is the precursor of intestinal metaplasia of ful patient follow-up, should further enhance knowledge of the esophagus. Both develop as a consequence of gastro- this disease.

Originally the term “Barrett’s esophagus” was used to HISTORY describe an esophagus in which a portion of the normal squamous mucosa was replaced by columnar epithelium. Stomach Versus Esophagus Since the 1950s, when the term became popularized, much Since the early 1900s, the presence of ulcers within a has been learned about this condition, and new concepts columnar epithelial-lined, tubular structure of the foregut continue to emerge that force reevaluation of this disorder. have been described. In the mid-1900s, the debate raged In fact, the meaning of the term Barrett’s esophagus is over whether these ulcers were located in a tubularized perhaps more ambiguous now than ever before. portion of stomach or in the esophagus. Using criteria “relevant to the physiologist, surgeon, and physician,” Dr. Norman Barrett, an influential British surgeon, wrote in 1950 that sections of the gastrointestinal tract are defined by Correspondence: Steven R. DeMeester, MD, Dept. of Cardiothoracic Sur- gery, Norris Comprehensive Cancer Center, 1441 Eastlake Ave., MS their mucosa, and therefore the esophagus is “that part of the 74, Los Angeles, CA 90033. foregut distal to the cricopharyngeal sphincter which is Accepted for publication November 4, 1999. lined by squamous epithelium.”1 He went on to state that 303 304 DeMeester and DeMeester Ann. Surg. ● March 2000 when ulcers are found below the squamocolumnar junction, Evolving Definition they represent gastric ulcers within “a pouch of stomach . . . The 3-cm Rule drawn up by scar tissue into the mediastinum” or more likely represent “examples of a congenital short esopha- In the late 1950s, the term Barrett’s esophagus was un- gus.” derstood to mean a columnar-lined esophagus, and this In 1953, Allison and Johnstone demonstrated that the rather vague definition persisted until the 1980s. Refine- tubularized portion of the foregut declared by Dr. Barrett to ments in the definition came about as a result of two issues. be stomach had no peritoneal covering, normal esophageal The first was that as endoscopy became more common, it musculature, and typical esophageal mucous glands. They was recognized that many patients had hiatal hernias and concluded that “it appears better to refer to that congenital esophagitis, and that it was difficult and error-prone for an abnormality which from the outside looks like esophagus endoscopist to determine precisely where the esophagus ended and the stomach began. In addition, Hayward in and from the inside looks like stomach as ‘esophagus lined 19616 had written that the “normal” esophagus could have with gastric mucous membrane.’ ”2 Although they called 1 to 2 cm of columnar mucosa at the distal end. Therefore, the mucosa “gastric,” they recognized that oxyntic cells to avoid an overdiagnosis of Barrett’s resulting either from were not present. failure to recognize a tubularized portion of herniated stom- In 1957, Dr. Barrett accepted the viewpoint that in ach on endoscopy, or from failure to allow for the “normal” some patients the columnar epithelium extended further 2 cm of columnar mucosa in the distal esophagus, a 3-cm proximally into the esophagus than could be accounted rule was introduced in the early 1980s. This rule stipulated for by a hiatal hernia. Further, he concurred with Allison that a diagnosis of Barrett’s esophagus required a minimum and Johnstone that the columnar mucosa, despite its of 3 cm of columnar mucosa to be present above the “gastric” appearance, did not contain oxyntic cells and perceived gastroesophageal junction. Although various au- did not function like gastric mucosa. He agreed to the thors used between 2 and 5 cm of columnar mucosa, some term “columnar-lined lower esophagus,” and subse- minimum length requirement was adopted by most physi- quently his name became synonymous with the condi- cians and persists in the minds of many people to this day. tion. Importance of Intestinal Metaplasia The second issue that forced refinement in the definition Congenital Versus Acquired of Barrett’s esophagus was the identification of a subset of patients with a columnar-lined esophagus who were at risk In their 1953 report, Allison and Johnstone had noted that for developing adenocarcinoma. In 1951, Bosher and Tay- the presence of a columnar-lined esophagus was usually lor7 described goblet cells indicative of intestinal metaplasia associated with a hiatal hernia, and that all of the patients within a columnar-lined esophagus. In subsequent publica- had reflux esophagitis. Still, like Dr. Barrett, they consid- tions Allison, Barrett, and others noted that the columnar ered it to be of congenital origin. Later Dr. Barrett, while mucosa of Barrett’s esophagus, although “gastric” in ap- 2,3 still favoring a congenital etiology, conceded that “if the pearance, was not normal gastric mucosa. Despite these cardiac valve of a normal person were to become incompe- observations, most physicians paid no particular attention to tent and if the lower esophagus were, as a result, to be the histology of the columnar lining within the esophagus. 8 bathed for a long time by digestive gastric juice, the squa- This changed in 1976 after a report by Paull et al described mous epithelium could be eaten away and totally replaced the results of manometrically guided biopsies at multiple levels from 11 patients with a columnar-lined esophagus. by columnar cells.”3 Each patient was found to have one or a combination of In 1959, Moersch et al4 reviewed the specimens from three histologic types of columnar epithelium: a gastric 36 patients who had undergone esophageal resections at fundic type composed of chief and parietal cells; a junc- the Mayo Clinic for esophagitis. They noted that “cells tional type composed of mucous glands without parietal resembling young columnar cells were seen occasionally, cells; and a specialized type with intestinal characteristics and thus the question of inflammatory metaplasia had to including villiform surface, mucous glands, Alcian blue- be considered.” This was the first convincing evidence staining goblet cells, and no parietal or chief cells. In these that the columnar epithelium of Barrett’s esophagus was 11 patients, Paull et al noted that the most prevalent type of acquired, and that it was caused by repetitive exposure of columnar epithelium was the specialized type, and when the distal esophagus to refluxed gastric juice. Subse- present, the specialized mucosa was always found adjacent quently several studies, including the 1970 landmark to the squamous epithelium in the most proximal portion of experimental report by Bremner et al,5 confirmed the the columnar epithelium. If the specialized mucosa did not association between a columnar-lined esophagus and gas- involve the entire columnar segment, then junctional mu- troesophageal reflux disease (GERD), and the congenital cosa was present below the specialized mucosa, and the theory was discarded. gastric fundic type of epithelium was found most distal. The Vol. 231 ● No. 3 Columnar Mucosa and Intestinal Metaplasia of the Esophagus 305 importance of the specialized or intestinal type of epithe- for Barrett’s. Using the traditional definition of at least 3 cm lium within a columnar-lined esophagus was hinted at in of columnar mucosa above the gastroesophageal junction, this report, in which 1 of the 11 patients was noted to have Barrett’s has been reported in 0.45% to 2.2% of all patients marked dysplastic changes within an 8-cm segment of spe- undergoing upper endoscopy and in up to 12% of patients cialized columnar-type epithelium. Subsequently, Haggitt undergoing endoscopy for symptoms of reflux.14 If all pa- suggested and Skinner and then Reid confirmed that the tients with a biopsy showing intestinal metaplasia, regard- intestinal type of columnar mucosa was premalignant, and less of length, were included in the definition, then the further was the only type associated with malignant degen- incidence increases to 9% to 32% of unselected patients eration.9–11 undergoing upper endoscopy.15 Based on an autopsy series, Cameron et al16 estimated the prevalence of traditional Traditional Definition of Barrett’s Barrett’s in the general population to be 376 cases per After these reports, the histology of the columnar-lined 100,000 population. Consequently, they suggested that for esophagus assumed great importance, and the definition of every known patient with Barrett’s, there might be 20 or Barrett’s esophagus evolved again to require both 3 cm of more unrecognized ones in the general population. It is columnar mucosa within the esophagus and intestinal meta- likely that the prevalence of any length of intestinal meta- plasia in histology. Consequently, a biopsy was now nec- plasia within the esophagus is even greater. essary to make a diagnosis of Barrett’s. Since the mid- 1980s, most physicians have restricted the term Barrett’s esophagus to patients meeting these criteria; however, it ETIOLOGY remains important even now to scrutinize the definition of Gastroesophageal Reflux Barrett’s esophagus used when reviewing any paper written on the subject. Typically, patients with Barrett’s esophagus have the classic gastroesophageal reflux symptoms of heartburn and Short-Segment Barrett’s and Intestinal regurgitation, with or without dysphagia. Often these pa- Metaplasia of the Cardia tients describe a longer duration of reflux symptoms than do Recently, with continued improvements in endoscopic patients with reflux disease without Barrett’s.17 Ambulatory equipment and the availability of potent acid-suppression pH monitoring demonstrates excessive acid reflux into the medication capable of healing esophagitis, it has become distal esophagus in nearly all patients with Barrett’s. The evident that short tongues of columnar mucosa can be found abnormal acid exposure is primarily due to a structurally extending up from a well-defined gastroesophageal junc- defective lower esophageal sphincter (LES), with defects in tion. Further, biopsies from a normal-appearing gastro- length, pressure, or both demonstrated manometrically in esophageal junction have been found to contain microscopic more than 90% of patients with Barrett’s.18 In addition, foci of intestinal metaplasia within cardiac mucosa, a con- patients with Barrett’s frequently have impaired esophageal dition referred to as intestinal metaplasia of the cardia. motility with low contraction amplitudes and an increased There is increasing evidence that, similar to traditional frequency of abnormal waveforms in the distal esophagus Barrett’s, these short lengths of intestinal metaplasia are compared with normal subjects.19,20 As a consequence, associated with gastroesophageal reflux and may have ma- refluxed material is poorly cleared from the esophagus. The lignant potential.12 combination of frequent reflux episodes and poor clearance allows prolonged bathing of the distal esophagus in refluxed gastric juice, resulting in severe mucosal injury. Current Definition In addition to an increased amount and exposure time of The relation between intestinal metaplasia present only at refluxed gastric juice, the composition of the refluxed juice the gastroesophageal junction, short-segment (Ͻ3 cm) Bar- likely contributes to the pathogenesis of Barrett’s esopha- rett’s, and traditional (Ն3 cm) Barrett’s remains controver- gus. The observations that gastric hypersecretion can be sial. However, an acceptable current definition of Barrett’s associated with Barrett’s and that Barrett’s has developed is a change from the normal esophageal squamous mucosa after total gastrectomy suggest that acid gastric juice is not to columnar mucosa of any length that is visible endoscop- the sole factor responsible for this disease.21 Recent reports ically and that on biopsy demonstrates intestinal metapla- are focusing on the importance of refluxed duodenal secre- sia.13 Using this definition, intestinal metaplasia confined to tions in the development of intestinal metaplasia. Patients the gastroesophageal junction would not be considered Bar- who reflux both gastric and duodenal juice have been found rett’s. to have a higher prevalence of esophagitis and Barrett’s esophagus than do patients who reflux gastric juice alone.18 Analysis of the composition of reflux in 281 patients with EPIDEMIOLOGY GERD demonstrated that the patients with the greatest The prevalence of Barrett’s esophagus is unknown, but degree of mucosal injury were more likely to have both any estimate depends significantly on the definition one uses gastroduodenal and acid reflux as opposed to pure gastric 306 DeMeester and DeMeester Ann. Surg. ● March 2000

Table 1. PREVALENCE OF INCREASED ESOPHAGEAL ACID EXPOSURE BY 24- HOUR pH MONITORING, INCREASED ESOPHAGEAL BILIRUBIN EXPOSURE BY BILITEC PROBE, AND STRUCTURALLY DEFECTIVE LES BY MOTILITY IN PATIENTS WITH INCREASING MUCOSAL INJURY SECONDARY TO GERD

Prevalence of Prevalence of Prevalence Increased Increased of Acid Bilirubin Defective Group Exposure (%) Exposure (%) LES (%) Figure 2. Dissociation curve for bile acids demonstrating the critical pH range from 3 to 6 where bile acids exist in their soluble, unionized form 1. No mucosal injury 35.4 30.0 51.5 and can penetrate cell membranes, accumulate within mucosal cells, 2. Esophagitis 80.0* 61.4* 85.7* and become toxic to the mitochondria. At pH 2, bile acids irreversibly 3. Short-segment 93.3* 73.3* 73.3* precipitate from solution, whereas at pH 7, bile acids exist in their Barrett’s noninjurious ionized form. 4. Long-segment 96.9‡ 84.4‡ 93.8† Barrett’s enter mucosal cells and cause injury, they must be soluble GERD, gastroesophageal reflux disorder; LES, lower esophageal sphincter. Data from reference 22. and unionized. At pH 7, greater than 90% of bile salts are in * P Ͻ .05 vs. group 1; ‡P Ͻ .05 vs. groups 1 and 2; ‡ P Ͻ .05 vs. groups 1 and solution and completely ionized. Acidification of bile to a 3. pH of less than 2 produces irreversible precipitation. Thus, under normal physiologic gastric conditions, bile acids pre- cipitate and are of minimal significance. However, in a more reflux (Table 1).22 In addition, patients with Barrett’s had a alkaline gastric environment, as can occur with the use of significantly higher prevalence and duration of abnormal acid-suppression medication, bile salts are partially dissoci- esophageal bilirubin exposure, which is a tag for duodenal ated. At a critical pH between 3 and 5, a mixture of ionized juice, compared with patients with only esophagitis.23 salt and the lipophilic, unionized acid is present. Unionized Among patients with Barrett’s, significantly greater esoph- bile salts can rapidly cross mucosal cell membranes. Once ageal bilirubin exposure has been demonstrated in those in inside the alkaline environment of the cell, they are con- whom dysplasia develops (Fig. 1). verted back into their ionized form. Consequently, they become trapped within the cell, accumulate, and ultimately become toxic to the mitochondria (Fig. 2). For bile salts to Role of Bile remain completely ionized and innocuous in a patient with Evidence is accumulating that bile salts are the noxious reflux disease taking acid-suppression medication, a gastric component in refluxed duodenal juice, and that their ability pH of 7 must be maintained 24 hours a day, 7 days a week 24 to cause cellular injury is pH-dependent. For bile salts to for the patient’s lifetime. Kuo and Castell have demon- strated that normal volunteers taking 20 mg omeprazole twice a day spend more than 30% of a 24-hour monitoring period with a gastric pH of less than 4. Thus, it is likely that such strict control of gastric acidity is not only impractical, but impossible unless very high doses of medication are used. Insufficient doses of medication may allow esopha- geal mucosal cell injury to occur while the patient remains relatively asymptomatic.25

Cellular Origin of Barrett’s It is generally accepted that Barrett’s esophagus is an acquired condition that occurs as a consequence of gastro- esophageal reflux. This implies that these patients at one time had normal squamous epithelium in their esophagus, Figure 1. Acid and bilirubin exposure times for patients with Barrett’s and that this epithelium subsequently was replaced. Al- esophagus with and without complications. The presence of a stricture, ulcer, low-grade dysplasia, or high-grade dysplasia was considered a though most clinicians accept this concept, the sequence by complication. The median is indicated by a line, the interquartile range which this occurs is controversial. Contained within the by a box. controversy are two issues: whether injured squamous mu- Vol. 231 ● No. 3 Columnar Mucosa and Intestinal Metaplasia of the Esophagus 307 cosa is first replaced by or transformed into columnar mu- cosa, which then undergoes a second transformation into intestinalized columnar mucosa; and whether columnar ep- ithelium and intestinalization gradually involve more of the esophagus, or whether Barrett’s rapidly develops to its full extent with little subsequent change. The answers to these questions are not known with certainty, but there is an ever-increasing body of data that provides insight about the process of esophageal columnarization and intestinalization.

Columnar-Lined Esophagus Step 1: Transition Zone From Esophageal Squamous Mucosa to Gastric Columnar Epithelium When considering the effects of acid reflux on squamous esophageal epithelium, a logical place to examine is the esophagogastric junction. Amazingly, despite years of in- terest and investigation, even the normal anatomy of this 6 area remains somewhat of an enigma. In 1961, Hayward Figure 3. Hayward’s depiction of the gastroesophageal junction and published a visionary treatise on this area that, although region of the cardia. Line X-Y crosses the esophagogastric junction. fascinating to read, is unfortunately devoid of any support- Everything above the line is esophagus, and everything below is stom- ach. From A to C is esophagus lined by squamous epithelium. The ing data. Careful review of the manuscript, however, sug- phrenoesophageal ligament inserts at B. From B to D is the area of the gests that his insight into this area was based on extensive esophagus called the cardia. It is partially lined by squamous epithelium dissection in probably many subjects. He began with a (from B to C) and partially lined by cardiac mucosa (from C to D). discussion of the meaning of the cardia—something equally Cardiac mucosa joins with fundic gastric mucosa at point E, whereas relevant today. He defined the cardia anatomically as the pure fundic mucosa is present at point F. (From Hayward J. The lower end of the esophagus. Thorax 1961; 16:36–41. Reproduced with per- lower portion of the esophagus between the insertion of the mission from the BMJ Publishing Group.) phrenoesophageal ligament and the gastroesophageal junc- tion (Fig. 3). The squamocolumnar junction, he noted, was usually located just distal to the insertion of the phrenoe- histologic differences between gastric mucosa in the fundus sophageal ligament, and consequently a portion of the car- or body of the stomach, and the mucosa immediately distal dia was “normally” lined by a short segment of columnar to the gastroesophageal junction. All three sites are charac- epithelium extending from the gastroesophageal junction terized by pure oxyntic mucosa. In contrast, the cardia, as proximally to the squamocolumnar junction. He called this described by Hayward, refers to the lower end of the esoph- columnar mucosa junctional epithelium, and noted that it agus in the region of the LES. It was from this cardia region, had the following characteristics: or the distal esophagus near the gastroesophageal junction, that Hayward identified the histologically distinct epithe- ● It was histologically distinct from normal gastric fundic and pyloric epithelium. lium that he called junctional mucosa. In the years since Hayward’s article, junctional mucosa has come to be called ● It did not secrete acid or pepsin but was resistant to both. cardiac mucosa. Therefore, it is important not to confuse the cardiac region of the stomach, which is located distal to the ● It was not congenital but acquired. gastroesophageal junction and has oxyntic mucosa, with the ● It was mobile and varied in length, creeping progres- sively higher into the esophagus with continued gastro- cardia, which is located at and proximal to the gastroesoph- esophageal reflux. ageal junction in the distal esophagus, and may be lined by cardiac (formally known as junctional) mucosa. ● It was potentially reversible with correction of reflux. Perhaps rather than focusing exclusively on the anatom- Hayward clearly placed the anatomic cardia and the lin- ical region, it would be best to combine the anatomic region ing junctional mucosa within the lower esophagus. Unfor- with the histology at that location to distinguish reliably tunately, there has been substantial confusion generated by between distal esophagus and stomach. An important and the terminology used to describe this area. The confusion unresolved issue is what is the “normal” mucosa at the centers around the fact that anatomists have divided the esophagogastric junction. Recall that Hayward stated that stomach into five parts, with the area around the esophago- “normally” the mucosa in this region was columnar, but gastric junction labeled as the cardiac portion of the stom- distinct from gastric fundic mucosa, and was not congenital ach. Despite the different names, there are no discernible but acquired. Perhaps by “normal” he meant common, be- 308 DeMeester and DeMeester Ann. Surg. ● March 2000 cause he recognized that it was not present at birth. Re- Therefore, it is our opinion that in the truly normal or cently, Chandrasoma26 conducted a review of the gastro- congenital state, squamous esophageal mucosa ends esophageal junction in a large number of autopsies in which abruptly at the gastroesophageal junction and abuts the there was no mention of GERD in the medical record. He oxyntic mucosa of the stomach. However, in many or per- determined that in most children and adults younger than 20 haps most older children and adults, chronic, low-level years of age, the squamous esophageal epithelium transi- reflux occurs, and repetitive bathing of the squamous mu- tioned directly with oxyntic mucosa of the gastric fundus, cosa at the gastroesophageal junction with gastric juice with no interposed segment of cardiac (junctional-type) induces the development of cardiac-type columnar mucosa. epithelium. Cardiac mucosa appeared in specimens from Recently, we reported that most of the patients being eval- those older than 20 years, but its length was almost always uated for symptoms of gastroesophageal reflux had cardiac less than 1 cm. From this he concluded that in the normal, mucosa juxtaposed between esophageal squamous and gas- nonrefluxing state, squamous esophageal epithelium tric fundic mucosa.28 Further, the cardiac epithelium almost changes abruptly to oxyntic mucosa of the gastric fundus. always had histologic evidence of inflammation unrelated to However, the presence of a small length of cardiac mucosa either Helicobacter pylori infection or mucosal pathology is common in adults. elsewhere in the stomach. Instead, the presence of inflamed Oxyntic mucosa, like that found in the stomach near the cardiac mucosa, or carditis, correlated closely with objec- gastroesophageal junction, is not affected by acid. Conse- tive markers of GERD, including an incompetent LES, quently, it is logical to conclude that cardiac mucosa, if not increased esophageal acid exposure on 24-hour pH moni- present at birth, must develop from injured squamous epi- toring, a hiatal hernia, and erosive esophagitis. We sug- thelium. This likely involves a change in the direction of gested that carditis may represent the earliest manifestation differentiation of the germinative squamous epithelial cells toward more acid-resistant columnar cells as a consequence of reflux disease, and that with continued reflux there might of repetitive exposure to acidic gastroesophageal reflux. In be a creeping columnarization of the squamous epithelium keeping with the concept that cardiac mucosa represents within the LES, a progressive loss of sphincter competency, transformed squamous mucosa, cardiac mucosa is the sim- and ultimately explosion of the disease into the body of the 28 plest type of columnar epithelium. It is characterized histo- esophagus. logically by the presence of only columnar cells and mucous Step 2: Extension of Columnar Mucosa cells. There are no specialized cells such as parietal cells, Proximally Into the Distal Esophagus chief cells, or goblet cells. The hypothesis that cardiac mucosa develops as a conse- Exposure of the squamous esophageal mucosa to gastric quence of acid-induced injury to esophageal squamous ep- juice in persons without prior surgery is most likely to occur ithelium is supported by both animal and clinical evidence. postprandially, when the stomach is distended and the distal Experimental evidence comes from a 1970 study by Brem- 29 5 esophageal sphincter is taken up by the enlarging fundus. ner et al. In this study a series of dogs underwent stripping Csendes et al30 have demonstrated that with increasing of the distal esophageal squamous mucosa with or without severity of gastroesophageal reflux, the squamocolumnar creation of a cardioplasty to destroy the function of the LES. junction progressively shifts or creeps proximally, resulting They noted extensive squamous cell reepithelialization in in a gradual increase in the length of cardiac-type columnar the animals without gastroesophageal reflux, whereas squa- mucosa within the esophagus. In addition, studies using mous regeneration was absent or minimal in the animals 24-hour pH monitoring have shown that as acid exposure with cardioplasty-induced acid reflux. In the animals with increases, the length of columnar mucosa within the esoph- reflux, the esophagus was reepithelialized by a columnar 31 epithelium that lacked submucosal glands and parietal agus steadily and significantly increases as well (Fig. 4). cells—the equivalent of cardiac mucosa in humans. Evi- The importance of this progressive increase in the length dence that columnar mucosa can replace normal esophageal of cardiac-type columnar mucosa in the esophagus with squamous epithelium in humans comes from follow-up increasing severity of GERD lies in the direct correlation studies in patients who have undergone partial esophago- between the length of cardiac mucosa and the likelihood of gastrectomy with an intrathoracic anastomosis of the esoph- finding an area of intestinal metaplasia. We found that the agus to the fundus of the stomach. This arrangement leads prevalence of any intestinalization within cardiac mucosa to a near-constant bathing of the remaining esophagus in increased from 12% when cardiac mucosa was limited to refluxed gastric juice, and over months to years columnar the gastroesophageal junction, to 50% with less than 3 cm, epithelium histologically identical to cardiac mucosa has and nearly 100% with 3 cm or more of cardiac mucosa (Fig. developed proximal to the anastomosis in what was unques- 5). Spechler et al32 reported similar findings. They noted tionably the esophagus, and in what unquestionably had that 36% of patients with a 1- to 2-cm columnar segment in been squamous epithelium. In a report by Lindahl et al,27 the esophagus had intestinal metaplasia, and this increased the mean time to detection of cardiac mucosa above the to 93% of patients when the columnar segment was 3 cm in anastomosis was 8.2 years (range 2.2–13.9 years). length or more. Vol. 231 ● No. 3 Columnar Mucosa and Intestinal Metaplasia of the Esophagus 309

small intestine marker proteins; however, in that study only three patients with cardiac mucosa were evaluated. Re- cently, Griffel et al34 have shown that the murine antibody DAS-1, which stains specialized columnar mucosa, reacts positively with cardiac mucosa, and that on repeat biopsies histologic evidence of intestinalization later developed in six of the seven patients. If true, these findings would suggest that biochemically cardiac mucosa and intestinal metaplasia are similar, and that cardiac mucosa is the pre- cursor of intestinalized columnar epithelium, or Barrett’s. Fitzgerald et al35 have recently suggested that the dynam- Figure 4. Correlation between acid exposure as determined by 24- ics of acid exposure may effect columnar cell proliferation hour pH monitoring and the length of cardiac-type columnar epithelium and differentiation. Using cultured human Barrett’s biopsy without intestinal metaplasia in the esophagus. The length of cardiac specimens, they demonstrated that continuous exposure to mucosa was measured from the endoscopic gastroesophageal junc- acidic media at pH 3.5 resulted in increased villin expres- tion to the site of the highest biopsy showing cardiac-type columnar sion (a marker for epithelial cell differentiation) and re- epithelium on histologic examination. duced cell proliferation. In contrast, villin expression was not detected when the culture medium was made more acidic (pH Ͻ 2.5). A dramatic increase in proliferation Step 3: Intestinalization of Cardiac Mucosa occurred when the Barrett’s tissue was exposed to a short The primary importance of inflamed cardiac mucosa is (1-hour) pulse of acidic medium (pH 3.5) followed by a that it seems to represent the only mucosal type that can return to neutral pH. Consequently, alterations in luminal progress to intestinal metaplasia.26 Intestinalization of the pH and the pattern of exposure may lead to altered growth cardiac mucosa occurs when hypertrophic cardiac mucous properties and may contribute to the molecular and struc- cells develop acid rather than neutral mucin and, most tural heterogeneity seen in the esophageal mucosa of pa- importantly, goblet cells appear. Once intestinalized, the tients with Barrett’s. cardiac mucosa seems to have an increased ability to with- stand damage by refluxed gastric juice, because there com- Intestinal Metaplasia of the Cardia monly is little or no inflammation present on biopsy. How- ever, the development of intestinal metaplasia within One unresolved issue is whether intestinal metaplasia cardiac mucosa is considered a detrimental or progressive limited to the region of the gastroesophageal junction de- change, because this mucosa is capable of further progres- velops in response to the same stimuli that produce tradi- sion to epithelial dysplasia and adenocarcinoma. Available tional, or long-segment, Barrett’s esophagus. Recently, we evidence would suggest that cardiac mucosa itself is benign, reviewed 411 patients with cardiac mucosa either at the and that it is only with the development of intestinal meta- plasia that the mucosa becomes premalignant.26 The specific cellular event that induces a change from cardiac mucosa to intestinalized cardiac mucosa is un- known. Current theories support a second insult, perhaps from noxious luminal contents. Data from our esophageal laboratory on patients with short lengths of columnar mu- cosa in the esophagus suggests that compared with patients without intestinal metaplasia, those with intestinal metapla- sia tend to have a longer duration of symptoms (5 vs. 10 years) and a greater frequency of abnormal exposure to refluxed bilirubin, as determined by Bilitec probe (Medtronic Functional Diagnostics, Shoreview, MN) mon- itoring for 24 hours. The frequency of abnormal acid expo- sure by 24-hour pH monitoring was similar between groups (DeMeester TR, unpublished data). It is also unclear whether the development of intestinal Figure 5. The prevalence of intestinal metaplasia in cardiac-type mu- metaplasia represents a phenotypic change secondary to the cosa of varying extents: limited to the gastroesophageal (GE) junction Ͻ Ն induction of genes, or a mutational event within the colum- and involving short ( 3 cm) and long ( 3 cm) segments of the esoph- 33 ageal body. (Reproduced with permission from Oberg S, DeMeester nar cells. Mendes de Almeida et al have demonstrated TR, Peters JH, et al. The extent of Barrett’s esophagus depends on the biochemically that both cardiac mucosa and intestinal meta- status of the lower esophageal sphincter and the degree of esophageal plasia express sucrase-isomaltase and crypt cell antigen, two acid exposure. J Thorac Cardiovasc Surg 1999; 117:572–580.) 310 DeMeester and DeMeester Ann. Surg. ● March 2000 gastroesophageal junction or extending up into the esopha- intestinal metaplasia. If CIM and Barrett’s represent two gus, and found that overall 35% of patients had intestinal unrelated conditions, each caused by different factors, then metaplasia. When only the patients with intestinal metapla- in some patients the two conditions should exist simulta- sia were compared, we noted that the length of intestinal neously—in other words, there should be some patients metaplasia was inversely correlated with LES pressure and with intestinal metaplasia at both the proximal and distal overall length and directly correlated with the percentage of ends of a segment of columnar epithelium, with cardiac time spent at less than pH 4 on 24-hour esophageal pH mucosa in between. This does not occur. Rather, patients monitoring.36 Our interpretation of this data is that both either have intestinal metaplasia throughout the entire intestinal metaplasia limited to the gastroesophageal junc- length of the columnar epithelium, or it is confined to the tion (CIM) and Barrett’s are related through the common proximal end of the columnar segment, and cardiac mucosa denominator of cardiac mucosa, and that CIM, Barrett’s, without goblet cells is found distally. and cardiac mucosa are all linked to gastroesophageal re- flux. Others, including Goldblum et al,37 have published Time Interval From Columnar to their opinion that intestinal metaplasia of the gastroesoph- Intestinalized Columnar Mucosa ageal junction is related to H. pylori infection and intestinal metaplasia in the stomach. One problem is that they failed There is increasing evidence, particularly from follow-up to distinguish clearly between gastric fundic and cardiac studies in children, that cardiac-type columnar mucosa de- mucosa, as evidenced by the histologic finding of parietal velops within the esophagus initially, and with time intes- cells on several “cardiac” biopsy specimens. Further, their tinalization may occur. In a series of 11 children aged 1.1 to study was flawed in that it lacked a reliable control group of 13.3 years (average 5.9 years) with symptoms of gastro- patients, because 7% of the nonreflux “control” patients esophageal reflux and a columnar-lined esophagus, Cooper were found to have esophagitis on endoscopy. Certainly H. et al42 found that only 1 child had intestinal metaplasia on pylori infection and intestinalization of the stomach are biopsy. In older children, Hassall et al43 and Hoeffel et al44 present in some patients with intestinal metaplasia of the reported the presence of intestinal metaplasia within a co- gastroesophageal junction. However, the intestinal metapla- lumnar-lined esophagus, and in two patients (ages 11 and sia related to H. pylori infection in these patients is often 16), adenocarcinoma was found within the Barrett’s. Qual- diffuse throughout the stomach. Intestinal metaplasia lim- man et al45 followed up 28 pediatric and 38 adult patients ited to the gastroesophageal junction does not appear related with a columnar-lined esophagus and noted that compared to either H. pylori infection or intestinal metaplasia of the with the pediatric patients, the percentage of adult patients stomach.28,38,39 with goblet cells and intestinal metaplasia was significantly increased. The youngest patient with goblet cells in their report was a 5-year-old with intestinal metaplasia found Location of Intestinal Metaplasia Within only at the gastroesophageal junction. They also described a Columnar-Lined Esophagus one 10-year-old patient with a columnar-lined esophagus Interestingly, as first described in the report by Paull et without intestinal metaplasia who was followed up with al,8 when intestinal metaplasia is found within cardiac mu- serial endoscopies and biopsies; at age 15, goblet cells cosa, it is always present at the most proximal portion of the developed in this patient. columnar mucosa at the squamocolumnar junction. Al- Another source for information about the time interval though intestinal metaplasia may involve the entire colum- from cardiac mucosa to intestinalized columnar epithelium nar segment, often an area of cardiac mucosa without in- are reports documenting follow-up on patients after esoph- testinal metaplasia is found distally near the gastro- agectomy with gastric pull-up reconstruction. In a review of esophageal junction. Morales et al40 and others have used 17 patients who had undergone partial esophagectomy with this finding to suggest that CIM and Barrett’s are not re- intrathoracic esophagogastrostomy, Hamilton and Yard- lated, or at least that CIM is not the precursor of Barrett’s.41 ley46 noted that columnar mucosa had developed above the Instead, we propose that this observation supports the con- esophagogastric anastomosis in 10 of the 17 patients. They cept that intestinalization of cardiac mucosa and prolifera- noted that cardiac-type columnar mucosa without goblet tion of Barrett’s occur as a consequence of the composition cells could be found as soon as 2 months after surgery in an and pH of the refluxed juice and the pulsatile or continuous area that had previously been histologically shown to be nature of the exposure. Extrapolating from the work of squamous epithelium. In two patients, intestinal metaplasia Fitzgerald et al,35 it is conceivable that because the proximal was found within the cardiac mucosa at the squamous portion of the columnar segment is the area furthest away junction at 76 and 106 months after surgery. from the stomach, it is the area most likely to be exposed in These findings would suggest that cardiac mucosa is the a more continuous manner to a pH of 3.5 or greater due to precursor of intestinal metaplasia, and that the process of the mixing of refluxed gastric juice and saliva. This would intestinalization occurs sequentially over a period of at least stimulate increased cellular differentiation and less prolif- 5 years. In contrast to this stepwise extension theory is the eration, creating the opportunity for the development of common belief that Barrett’s develops rapidly and to its full Vol. 231 ● No. 3 Columnar Mucosa and Intestinal Metaplasia of the Esophagus 311 extent with little subsequent change. This concept is based invasive adenocarcinoma. Cameron and Carpenter53 have largely on a study by Cameron et al47 in which 21 patients demonstrated that areas of dysplasia and cancer within long with 3 cm or more of intestinal metaplasia were followed up segments of intestinal metaplasia are often small and for a mean of 7.3 years. At the end of follow-up, the length patchy, and that microscopic areas of different grades of of Barrett’s was not significantly different from what it had dysplasia are often dispersed throughout the Barrett’s. They been initially, and the mean length of Barrett’s was similar suggested that dysplasia develops simultaneously in many in all age groups. However, we contend that patients with areas and ultimately becomes confluent rather than spread- long segments of Barrett’s uniformly have profound defects ing progressively outward from one site. In addition, they in LES function, and further deterioration would not be noted that cancers developed throughout the length of the expected. Consequently, the fact that the length of Barrett’s intestinal metaplasia, including distally near the stomach. did not change in their study is not surprising. Important This is in contrast to prior reports indicating that the cancers future studies documenting the natural history of short- always occur near the squamocolumnar junction at the segment Barrett’s and intestinal metaplasia limited to the proximal extent of the Barrett’s. This is likely a conse- cardia should help clarify this issue. quence of imprecision in the use of the term Barrett’s. As noted above, if intestinal metaplasia is present in a segment of columnar epithelium, it is always at the proximal end of METAPLASIA, DYSPLASIA, the columnar segment. It may extend distally to involve CARCINOMA SEQUENCE most or all of the columnar segment, but it may be limited The development of intestinal metaplasia within cardiac to the area near the squamocolumnar junction. Because mucosa heralds the onset of a mucosal change that ulti- available evidence suggests that adenocarcinoma occurs mately may lead to the development of esophageal adeno- only within areas of intestinal metaplasia, it is likely that in carcinoma. Consequently, the term “benign Barrett’s” is an these prior reports the intestinal metaplasia was limited to oxymoron. With continued inflammation and irritation of the proximal portion of the columnar-lined distal esopha- the metaplastic intestinal epithelium, some patients will gus. progress through low-grade dysplasia, high-grade dysplasia, and subsequently invasive adenocarcinoma. It is unknown DIAGNOSIS whether the movement toward cancer is due to mitogenesis secondary to chronic mucosal injury, or mutagenesis as a When performing an esophagogastroduodenoscopy, the consequence of exposure to a mutagen. One theory is that endoscopist should carefully note the location of the gas- bile salts in their unionized state act as mutagens, and troesophageal and squamocolumnar junctions as well as the studies are underway to investigate this possibility. In our crura. Barrett’s esophagus should be suspected when the laboratory, cell culture studies indicate that repetitive short gastroesophageal junction is located distal to the squamo- exposure of cells to bile salts at a pH of 5 to 7 increases the columnar junction, or when there are pink tongues of gas- mutation frequency without altering the growth curve of the tric-like mucosa extending up into the esophagus above an cultured cells (DeMeester TR, unpublished data). If bile aligned gastroesophageal and squamocolumnar junction. salts are demonstrated to contribute to the development of Precise determination of the gastroesophageal junction can malignancy, then early intervention with an antireflux pro- be difficult in these patients because frequently a hiatal cedure should be encouraged. The precise risk of develop- hernia is present, and as a group patients with Barrett’s have ing adenocarcinoma in patients with Barrett’s esophagus is a very low LES pressure.13 Identification of the proximal unknown; however, it likely is 0.2% to 2.1% per year for a margin of the gastric folds with the stomach and esophagus patient without dysplasia, which is an incidence 30 to 125 deflated facilitates localization of the gastroesophageal times that of the general population.48 A common estimate junction.54 is that there will be one cancer found for each 100 patient- Because both erosive esophagitis and erythema of the years of follow-up. This increased incidence of cancer is the esophagus may be confused visually with Barrett’s, confir- rationale for enrolling all patients with Barrett’s esophagus mation of the diagnosis requires histologic identification of in a surveillance program, and heightening the surveillance goblet cells within a columnar mucosa devoid of parietal in patients with low-grade dysplasia. cells. To maximize the likelihood of identifying intestinal Although intestinal metaplasia is itself a premalignant metaplasia, multiple biopsy samples should be taken at the condition, the development of high-grade dysplasia is asso- proximal extent of the columnar mucosa, just below the ciated with a significantly increased risk for adenocarci- squamocolumnar junction, where intestinal metaplasia, if noma. Several reports have correlated the progression of present, will always be found.8 If there is endoscopic sus- dysplasia within Barrett’s with other cellular and genomic picion of Barrett’s, the endoscopist should ascertain the alterations, including mutations of the tumor suppressor length of involved esophagus and carefully examine the gene p53, aneuploidy, and loss of the Y chromosome.49–52 mucosa for any suspicious lesions. Biopsies of all mucosal Importantly, any focus of intestinal metaplasia is capable of abnormalities should be performed; in addition, once a undergoing dysplastic change and ultimately becoming an diagnosis of Barrett’s is established, biopsies should be 312 DeMeester and DeMeester Ann. Surg. ● March 2000 obtained at 1- to 2-cm intervals throughout the length of the Table 2. COMPARISON OF COST OF metaplastic mucosa to evaluate for the presence of dyspla- SURVEILLANCE MAMMOGRAPHY FOR sia. BREAST CANCER TO COST OF Patients with intestinal metaplasia limited to the cardia do SURVEILLANCE ENDOSCOPY FOR not have an endoscopically visible area of columnar epithe- PATIENTS WITH BARRETT’S ESOPHAGUS lium within the esophagus. Thus, diagnosing CIM requires that biopsy samples be taken from a normal-appearing gas- Breast Cancer/ Barrett’s Esophagus/ troesophageal junction. A combination of circumferential Mammography EGD antegrade biopsies as well as biopsies taken with the esoph- Cost per cancer $54,513 $37,928 agoscope retroflexed in the stomach will most reliably iden- detected tify CIM. These biopsies, in our opinion, should be a Cost per cure $83,292 $83,340 standard part of the endoscopic evaluation of all patients Cost per life-year saved $57,926 $ 4,151 suspected of having GERD. EGD, esophagogastroduodenoscopy. Data from reference 57. TREATMENT Surveillance Endoscopy in patients with Barrett’s is based on available data about The rationale for endoscopic surveillance in patients with the likelihood and rapidity of progression. Important factors Barrett’s is straightforward: to detect progression of disease include the length of intestinal metaplasia and whether toward cancer, and to allow early intervention while cure is dysplasia is present. In longer segments of intestinal meta- still likely. Predicated on this rationale is the concept that plasia, there is more mucosa at risk for progression, and the surveillance will be frequent enough that rarely, if ever, patients with low-grade dysplasia have already moved a does the disease progress beyond an early, curable stage step closer to cancer. We advise our patients with any length during the surveillance interval, and that once disease pro- of intestinal metaplasia within the esophagus to have yearly gression is detected, an intervention is performed. Contin- endoscopy, and we increase the frequency to every 3 to 6 ued surveillance despite disease progression may help clar- months for patients with dysplasia. Sampliner and the Prac- ify the natural history of the disease but does little to benefit tice Parameters Committee of the American College of the patient. Controversy exists, however, about the point at Gastroenterology13 have recently suggested that surveil- which there has been sufficient progression to warrant in- lance may be extended to every 2 to 3 years in Barrett’s tervention, and what intervention should be done. patients lacking dysplasia on systematic biopsy at two en- Provenzale et al48 used a Markov model of decision doscopies. For patients with low-grade dysplasia, they rec- analysis to demonstrate that aggressive endoscopic surveil- ommend that surveillance endoscopy be performed at lance combined with esophagectomy for high-grade dyspla- 6-month intervals for the first year, and subsequently done sia would markedly reduce the incidence of cancer in pa- yearly if there has been no change. tients with Barrett’s esophagus and would increase both overall and quality-adjusted life expectancy. The predic- Medical Therapy of Barrett’s tions of the model have now been verified in several clinical series. We and others have reported that patients with ade- There are three goals for treating patients with Barrett’s nocarcinoma detected within a surveillance program present esophagus: stop reflux, promote or induce healing or regres- at an earlier stage and have a significantly improved long- sion of the metaplastic epithelium such that the risk mucosa term survival after esophagectomy compared with patients (intestinal metaplasia) is eliminated, and halt progression to who present de novo with symptoms.55,56 dysplasia and cancer. Most patients with Barrett’s are The cost effectiveness of surveillance endoscopy for Bar- treated medically; however, adequate medical therapy is rett’s has recently been compared with mammography for difficult because of the degree of impairment of the LES and the detection of breast cancer (Table 2).57 The cost per the poor esophageal body motility frequently present in cancer detected and the cost per patient cured were similar, patients with Barrett’s. This is likely the reason why the but the cost per life-year saved was dramatically lower for least controlled symptom in patients with Barrett’s receiv- surveillance endoscopy in Barrett’s patients than mammog- ing medical treatment is regurgitation.58 Medical treatment raphy in women. This likely is because outside of a surveil- options are limited to dietary and lifestyle modifications, lance program, esophageal cancer often presents in an ad- promotility agents, and acid-suppression therapy. Recently, vanced stage, and the clinical course is rapidly fatal. In Sampliner and the Practice Parameters Committee of the contrast, the lag time between mammographic and palpable American College of Gastroenterology13 stated that “the detection of breast cancer is shorter, and the clinical course goal of therapy of Barrett’s esophagus should be the control with breast cancer is often protracted. of the symptoms of GERD” and that “symptom relief is an The recommended frequency for endoscopic surveillance appropriate endpoint for the therapy of Barrett’s esopha- Vol. 231 ● No. 3 Columnar Mucosa and Intestinal Metaplasia of the Esophagus 313 gus.” However, this viewpoint flies in the face of logic. whom heartburn, regurgitation, or both occurred at least Gastroesophageal reflux causes both Barrett’s and esopha- once a week compared with persons without these symp- geal cancer. Symptoms are not part of the pathophysiology toms. Interestingly, they noted that the risk of esophageal of the disease; rather, they are merely the variably expressed adenocarcinoma was three times higher among patients who byproduct of reflux. Many patients with Barrett’s have few used medication for symptoms of reflux compared with or no reflux symptoms, probably as a consequence of an those who did not use acid suppression medications. Others, altered sensitivity of the metaplastic epithelium to refluxed including Ortiz et al62 and Hameeteman et al,66 have also acid. Consequently, the eradication of symptoms, if present, linked medical therapy for Barrett’s esophagus with pro- cannot be equated with elimination of reflux. Katzka and gression to dysplasia and adenocarcinoma. In the study by Castell59 demonstrated that standard-dose omeprazole (20 Hameeteman et al from the Netherlands, 50 patients with a mg/day) fails to suppress acid sufficiently to keep gastric pH columnar-lined esophagus were treated medically and fol- neutral for a full 24 hours in patients with Barrett’s. Further, lowed up from 1.5 to 14 years (mean 5.2 years). Of these 50 increasing the dose of omeprazole until all symptoms were patients, initially only 34 had intestinal metaplasia on bi- alleviated was an unreliable measure of effective therapy, opsy of the columnar mucosa. At the completion of the because 80% of patients studied with 24-hour pH monitor- study, 37 patients had intestinal metaplasia, indicating that ing still had abnormal distal esophageal acid exposure. Barrett’s developed in 3 patients during the 5-year study Sampliner likewise found that high-dose proton pump in- period. In addition, at the start of the study, six patients had hibitor administration (lanisoprazole, 60 mg/day) failed to low-grade dysplasia and one patient had high-grade dyspla- normalize the 24-hour pH test in more than one third of sia. By the end of the 5-year study period, 10 patients had patients with Barrett’s tested while receiving therapy.58 low-grade dysplasia, 3 had high-grade dysplasia, and 5 had Consequently, for medical therapy to achieve the first of adenocarcinoma.66 Similarly, Sharma et al67 followed 32 the three goals of treatment for patients with Barrett’s, medically treated patients with short-segment Barrett’s prevention of reflux, the medical therapy will have to be (mean length 1.5 cm) for a mean of 36.9 months and found aggressive and not guided by symptoms. Most patients a 5.7% annual incidence of progression to dysplasia. During require a minimum of twice-daily omeprazole. All patients the 98 patient-years of follow-up in their series, high-grade should be tested while receiving acid-suppression therapy dysplasia developed in two patients, and one of these pa- with 24-hour pH monitoring, and as suggested by Castell et tients progressed to cancer. Recall that the expected rate of al,60 the majority are likely to demonstrate night-time acid cancer is 1 per 100 patient-years of follow-up. All patients breakthrough and will need an evening dose of an H2 in the study by Sharma et al were treated with omeprazole, receptor antagonist in addition to the twice-daily proton ranitidine, and/or promotility agents. They commented that pump inhibitor. Lastly, given the concern about creating most patients developed dysplasia while taking acid-sup- unionized, lipophilic bile salts, it may be that the endpoint pression medication, and they concluded that medical treat- of acid suppression in these patients needs to be a gastric pH ment does not prevent the development of dysplasia. of 7 rather than just a pH of more than 4.25 In the final analysis, incomplete therapy may prove to be worse than no Antireflux Surgery for Barrett’s therapy. The second and third goals of therapy in patients with Some of the problems and concerns associated with med- Barrett’s are to eliminate the risk mucosa (i.e., intestinal ical therapy for Barrett’s include issues about patient com- metaplasia) and to prevent progression to dysplasia and pliance, the safety of long-term high-dose acid-suppression cancer. Medical therapy has not been shown to achieve therapy, and the prompt return of symptoms after drug either of these goals reliably. Several reports have con- withdrawal. In contrast, antireflux surgery restores LES cluded that medical therapy does not cause regression of function and abolishes reflux of gastric contents into the intestinal metaplasia.61–63 This may be different in patients esophagus. Consequently, an antireflux operation ends the with short-segment Barrett’s. Weston et al64 have described repetitive injury to both the metaplastic and normal esoph- the loss of goblet cells from lengths of intestinal metaplasia ageal mucosa. Randomized clinical studies have confirmed less than 2 cm in 32% of patients treated medically for 1 to superior control of reflux after antireflux surgery compared 3 years. In contrast, only 2 of 29 patients (7%) with lengths with medical therapy, and antireflux surgery has been of intestinal metaplasia 3 cm or more had loss of goblet proven safe, effective, and durable.62,68 In addition, many cells. patients are candidates for a minimally invasive laparo- With respect to the efficacy of medical therapy in pre- scopic approach associated with a short hospital stay and venting progression of Barrett’s to dysplasia and cancer, rapid recovery. We therefore favor the performance of an there is speculation that prolonged and perhaps inadequate antireflux procedure in patients with Barrett’s. acid suppression may actually promote the development of There have been conflicting reports about whether intes- Barrett’s and the complications of Barrett’s.25 Lagergren et tinal metaplasia regresses after antireflux surgery. Brand et al65 recently reported that the risk of esophageal adenocar- al, in 1980,69 described complete regression in 4 of 10 cinoma was increased nearly eightfold among persons in patients with Barrett’s who underwent fundoplication. Sub- 314 DeMeester and DeMeester Ann. Surg. ● March 2000

compared with two medically treated patients. They con- cluded that compared with medical therapy, an antireflux operation in patients with Barrett’s was significantly asso- ciated with the prevention of dysplasia and cancer. Simi- larly, Katz et al81 followed 102 patients with Barrett’s for a mean of 4.8 years. By 3 years, dysplasia had developed in approximately 8% of the medically treated patients. In con- trast, patients treated by antireflux surgery had a signifi- cantly reduced risk of developing dysplasia (P ϭ .03). One factor complicating any analysis of progression of Barrett’s after antireflux surgery is that the cellular and genetic alterations leading to the development of dysplasia Figure 6. Review of the English language literature identified 11 series and adenocarcinoma may have already occurred before per- in which patients with Barrett’s were followed up after antireflux surgery. formance of the antireflux procedure. It has been estimated A total of 340 patients were followed up for an overall mean of 4.4 years (range 1–18 years). Most patients (74%) did not have any change in their to take up to 6 years for adenocarcinoma to develop within Barrett’s with follow-up. Regression occurred in 17% of patients, pro- Barrett’s with low-grade dysplasia, and thus some cancers, gression in 9%. particularly those that present during the first few postop- erative years, probably do not represent progression of sequently, most reports have demonstrated that although disease after surgery. McDonald et al77 made this point in a some regression of the length of Barrett’s is common, study from the Mayo Clinic. They found invasive adeno- complete regression occurs only rarely. Review of the En- carcinoma in two patients and carcinoma in situ in one glish language literature since 1977 documents follow-up patient during surveillance after antireflux surgery, but they on 340 patients after antireflux surgery.10,62,69–77 Complete noted that carcinoma did not develop in any patient after 39 regression occurred in only 13 patients; in 256 of the 340 months despite a median follow-up of 6.5 years and a patients (74%), the Barrett’s epithelium remained un- maximum follow-up of 18.2 years. changed (Fig. 6). Thus, the available literature would sug- Review of the English language literature since 1975 gest that regression of traditional Barrett’s cannot reliably revealed 11 series and a total of 346 patients with Barrett’s be predicted or anticipated after antireflux surgery. followed after fundoplication.10,62,69–71,73–75,77,80,82 Pa- In contrast to the unreliable regression of traditional or tients were found to have esophageal adenocarcinoma after 3-cm segments of Barrett’s, we recently demonstrated that antireflux surgery in only 7 of the 11 reports. Apart from 73% of patients with intestinal metaplasia at the gastro- these series, four isolated reports were found describing esophageal junction had complete regression after antireflux adenocarcinoma developing in patients with Barrett’s after surgery.78 By comparison, only 4% of patients with an an antireflux operation.9,83–85 Although the length of fol- endoscopically visible segment of Barrett’s had loss of low-up was not always available, 11 of the 19 cancers intestinal metaplasia after an antireflux procedure. Many of (58%) developed within 3 years of fundoplication, and 15 these patients demonstrated the development of squamous (79%) developed within 5 years of fundoplication. The islands or a shorter length of Barrett’s, but in our opinion remaining four cancers developed from 5 to 10 years after regression should be considered only when there has been fundoplication, but in each case the patient had recurrent complete loss of the mucosa at risk for malignant degener- reflux on the basis of symptoms or positive 24-hour pH ation. Recently, Low et al79 also reported complete regres- monitoring (Fig. 7). Thus, a functioning fundoplication sion of Barrett’s esophagus after antireflux surgery. In their seems to provide protection from progression of Barrett’s to report, loss of intestinal metaplasia occurred in 2 of 14 adenocarcinoma. patients followed up for a mean of 25 months after surgery. Perhaps of greater importance is the issue of progression Mucosal Ablation of Barrett’s to dysplasia or cancer after surgical treatment of reflux disease. Compared with medical therapy, antireflux Theoretically, the ideal treatment for a patient with Bar- surgery is associated with a reduced incidence of dysplasia rett’s esophagus is one that will restore the normal squa- and adenocarcinoma. McCallum et al80 prospectively fol- mous mucosa and eliminate the cancer risk associated with lowed 181 patients with Barrett’s. Twenty-nine had antire- intestinal metaplasia. Currently, neither medical nor surgi- flux surgery; the remaining 152 were treated medically. cal therapy reliably offers this in patients with lengths of After a mean follow-up of 62 months in the surgical group intestinal metaplasia of 3 cm or more. However, several and 49 months in the medical group, there was a significant experimental techniques for mucosal ablation show prom- difference in the incidence of dysplasia and adenocarci- ise. Investigational techniques include the use of laser, noma. Dysplasia was found in 3.4% of the surgical group photodynamic therapy (PDT), multipolar electrocoagula- and 19.7% in the medical group. No patient in the surgically tion, and the ultrasonic aspirator. In animal and limited treated group developed adenocarcinoma of the esophagus, clinical trials, some of these techniques have successfully Vol. 231 ● No. 3 Columnar Mucosa and Intestinal Metaplasia of the Esophagus 315

tastases were found. Three patients died during treatment (3%). Biopsies showed no evidence of subsquamous glan- dular mucosa in 98% of patients; however, in 2% of patients areas of Barrett’s with high-grade dysplasia were found below squamous mucosa, and in one patient a subsquamous adenocarcinoma was found 6 months after PDT treatment. The most promising ablation technique uses the Cavitron Ultrasonic Surgical Aspirator (CUSA; Valley Lab, Boulder, CO) device to remove the mucosa without violating the muscularis mucosa. Animal studies have demonstrated that strictures are not produced if the muscularis mucosa is kept intact.88 In addition, the entire area of intestinal metaplasia can be eradicated at a single setting, and rather than de- Figure 7. Review of the English language literature identified 11 series stroying the mucosal cells, with this technique the aspirated in which patients with Barrett’s were followed up after antireflux surgery. cells are available for cytologic review. When combined In 4 of the 11 series, adenocarcinoma developed in no patients during the follow-up period. A total of 346 patients were followed up and 12 with an antireflux procedure, this technique is likely to be cancers occurred. Seven additional cases of adenocarcinoma devel- the most reliable method to ablate Barrett’s and allow oping in Barrett’s esophagus after an antireflux procedure were found in squamous regeneration. Refinements in the different tech- other reports. Thus, a total of 19 adenocarcinomas that occurred after niques and careful clinical evaluation of the completeness of an antireflux procedure were found in the literature. Each case of cancer ablation, long-term complication rates, and impact on pro- is plotted on the time line at the point where it occurred. Time is marked in 5-year segments, with 0 the time of the antireflux procedure. Note the gression of Barrett’s to dysplasia and cancer with each of maximum as well as the mean or median follow-up for each of the 11 the techniques should be forthcoming in the near future. series. Despite the long follow-up in these series, most cancers are clustered between 0 and 5 years after the antireflux procedure. These cancers probably occurred as a consequence of cellular and genetic Barrett’s With Low-Grade Dysplasia alterations that took place before the fundoplication. For each of the cancers occurring after 5 years, there was evidence either by the recur- Identification of low-grade dysplasia on biopsies from a rence of reflux symptoms or a positive 24-hour pH test that the fundo- patient with Barrett’s is of concern because it may represent plication had failed. progression of disease along a continuum that ends with adenocarcinoma of the esophagus. Patients who are found to have low-grade dysplasia on the basis of a random biopsy produced ablation of the columnar mucosa and subsequent should undergo repeat endoscopy with systematic four- squamous reepithelialization. quadrant biopsies every 1 to 2 cm throughout the length of Elimination of gastroesophageal reflux is critical to suc- the Barrett’s segment to rule out high-grade dysplasia or cessful squamous reepithelialization after mucosal ablation cancer. In addition, any abnormal-appearing area within the of Barrett’s. Persistent or recurrent areas of intestinal meta- columnar segment should be carefully biopsied. Because of plasia plague most series of mucosal ablation in patients the potential for confusion between inflammatory atypia and taking proton pump inhibitors. In contrast, Salo et al86 low-grade dysplasia, patients who have not been treated for recently reported successful ablation of Barrett’s using Nd: reflux should receive a 3-month course of intensive acid- YAG laser after antireflux surgery. They followed up 11 suppression therapy with 40 mg omeprazole twice a day. patients for a mean of 26 months after the last laser treat- These patients should then undergo repeat endoscopy and ment and noted complete squamous regeneration in all biopsy. If low-grade dysplasia is again identified, these patients. There was no residual metaplastic epithelium ex- patients should undergo an antireflux operation followed by cept in two patients with persistent intestinal metaplasia at endoscopic surveillance. In our series of 60 patients with the gastroesophageal junction. intestinal metaplasia of the esophagus or esophagogastric One problem with ablation for Barrett’s is that in most junction, we found that preoperative low-grade dysplasia patients the intestinal metaplasia is circumferential, and was present in 10 patients. In 7 of the 10, the low-grade often involves a large surface area. Many of the ablation dysplasia reverted to Barrett’s without dysplasia after an techniques suffer from an imprecise control of the depth of antireflux procedure.78 Similarly, Low et al79 noted that 4 of injury, and when applied circumferentially have a high rate their 14 patients had low-grade dysplasia, and in all 4 of stricture formation. In a review by Overholt et al,87 patients it regressed to Barrett’s without dysplasia after esophageal strictures developed in 34% of patients treated surgery. If the low-grade dysplasia persists after antireflux with PDT. PDT eradicated Barrett’s in 8% of patients, but surgery, consideration should be given to mucosal ablation this increased to 43% with the addition of Nd:YAG laser with the ultrasonic aspirator. Further, patients with low- ablation. In 23% of patients treated with PDT for adenocar- grade dysplasia should be under frequent surveillance, be- cinoma in Barrett’s the malignancy persisted, and in both cause progression to high-grade dysplasia and cancer can patients who underwent surgical resection, lymph node me- occur quickly. 316 DeMeester and DeMeester Ann. Surg. ● March 2000

Barrett’s With High-Grade Dysplasia The diagnosis of Barrett’s, dysplasia, and cancer contin- ues to rely on histologic evaluation of biopsy specimens. Optimal treatment for patients with Barrett’s and high- Most endoscopic biopsies are obtained randomly from grade dysplasia remains controversial. The major issues are: within Barrett’s in the sense that unless an ulcer or nodule 1. What is the likelihood that adenocarcinoma will de- is noted, the biopsies are taken from an area of Barrett’s that velop in a patient with high-grade dysplasia, and over looks like any other area. It is important that the biopsies are what time interval does this typically occur? taken circumferentially at close intervals, but the impor- 2. What is our ability with current technology to detect tance of using “jumbo” forceps and the number of biopsies a focus of adenocarcinoma within Barrett’s esopha- that should be taken per linear 2 cm of esophagus continue gus with high-grade dysplasia—and the corollary, to be debated. Given the typically large surface area of how likely is it that no cancer is present if the Barrett’s, as well as the irregular distribution of dysplastic endoscopy and biopsies show no cancer? changes, there is an inherent possibility of sampling error. 3. What are the treatment options for high-grade dys- Certainly taking multiple biopsy samples will reduce the plasia, and what outcomes are associated with each? error rate; however, the only way to eliminate the possibility of sampling error is to remove the entire mucosa. This fact Issue 1 is reinforced by data suggesting that up to 60% of patients Progression to adenocarcinoma was reported to occur in who have an esophagectomy for high-grade dysplasia are 26% of 58 patients with high-grade dysplasia followed for a found on careful pathologic examination of the surgical median of 24 months by Levine et al from Seattle.89 How- specimen to have an unidentified focus of carcinoma despite ever, they excluded 12 patients found to have adenocarci- numerous preoperative biopsies that showed only high- noma on an early rebiopsy after initially diagnosing only grade dysplasia.55,96 high-grade dysplasia. Inclusion of these patients would in- Endoscopic ultrasound has been used in patients with crease the frequency of progression to 39% in this group of high-grade dysplasia in an effort to detect areas of cancer. patients followed up for 2 years. Similarly, Schnell et al90 However, current ultrasound probes, especially the balloon- reported progression to adenocarcinoma in 8 of 42 patients tipped probes, cannot provide sufficient detail to evaluate (19%) with high-grade dysplasia. The median length of the esophageal wall superficial to the muscularis propria. follow-up was not specified; however, it was noted that five Cameron and Carpenter53 reported that endoscopic ultra- of the eight patients who progressed to carcinoma did so sound identified only one of four early invasive cancers and within 1 year. The literature would suggest that although the gave one false-positive result in nine patients without can- time required to progress from high-grade dysplasia to cer. In the future, it is hoped that new high-frequency cancer is variable, most cancers develop within 3 years, and ultrasound probes will be able to assess with accuracy often the cancer is found within several months of initially cancers limited to the mucosa and submucosa. Recently, finding high-grade dysplasia. Williamson et al91 reported endoscopic fluorescence techniques have been used to try to that five patients with Barrett’s and high-grade dysplasia identify areas of high-grade dysplasia or cancer within progressed to cancer at intervals of 2, 2, 5, 6, and 8 months Barrett’s based on the different autofluorescence spectra of after developing high-grade dysplasia. normal versus metaplastic versus dysplastic tissue.97,98 The utility of these techniques remains to be determined. Issue 2 Issue 3 Currently, the only way to detect a focus of adenocarci- noma within Barrett’s esophagus is by histologic examina- Treatment options for patients with Barrett’s and high- tion of a biopsy specimen, but differentiation between high- grade dysplasia include continued endoscopic follow-up, grade dysplasia and carcinoma is difficult and requires an some form of mucosal ablation, or resection of the esoph- experienced pathologist. Efforts are underway to identify agus. The main issue to focus on during the next several markers that can indicate which patients with high-grade years is the outcome associated with the various treatments dysplasia are likely to progress rapidly to cancer. Examples for high-grade dysplasia. Important outcome measures in- of studied markers include DNA content abnormalities such clude the recurrence rate for intestinal metaplasia, dysplasia, as increased G2/tetraploid fractions, aneuploid cell popula- and cancer, treatment-related complication and death rates, tions, or both; mutations of the p53 tumor suppressor gene; and patients’ satisfaction with lifestyle and ability to eat. As c-erbB-2 oncogene overexpression; increased expression of more data about these issues become available, physicians epidermal growth factor receptor and transforming growth and patients will have a better understanding of the pros and factor-alpha; decreased activity of glutathione S-transferase; cons for each treatment option. p16 promoter hypermethylation; and increased expression Currently one option is to repeat endoscopy and biopsy in of inducible nitric oxide synthase and cyclooxygenase- patients with high-grade dysplasia until a definitive diagno- 2.92–95 As yet, no clear correlation has been established that sis of cancer is established. The obvious advantage of this allows any of these to be clinically useful in assigning risk type of approach is that the patient is spared the potential of progression to an individual patient. risks of surgery until cancer is identified. This is balanced Vol. 231 ● No. 3 Columnar Mucosa and Intestinal Metaplasia of the Esophagus 317 by the need for repeat endoscopies with multiple biopsies on the English literature from 1990 to 1996 and analyzed the a frequent basis, and the real possibility of a sampling error outcome from 10 series totaling 110 patients who under- resulting in a missed focus of adenocarcinoma. An adeno- went esophagectomy for high-grade dysplasia. The death carcinoma that is missed and allowed to invade into the rate was 2.6%, and the overall 5-year survival rate was 82%. submucosa has a 50% likelihood of associated lymph node All patients were thought before surgery to have only high- metastases.99 Unfortunately, in these patients the opportu- grade dysplasia; however, cancer was found in 56 of the 110 nity for cure will have been compromised and the benefits patients (51%), and in 15 patients the tumors were stage IIa of endoscopic surveillance potentially negated. or higher. This concern is not theoretical. In the study by Hameete- We have recently reviewed our experience with high- man et al,66 five patients were found during surveillance to grade dysplasia at the University of Southern California.100 progress to adenocarcinoma, and three underwent resection. Repeat biopsy in 23 patients referred for high-grade dyspla- All three had cancers that invaded into the muscularis sia demonstrated a focus of adenocarcinoma in 9 patients propria. These are relatively advanced tumors, and our data (39%). After esophagectomy, review of the specimen dem- would suggest that 80% of tumors penetrating this depth onstrated an additional six cancers. Thus, of 23 patients into the wall of the esophagus will have lymph node me- referred for high-grade dysplasia, ultimately cancer was tastases.99 Likewise, in the series by Williamson et al,91 five found in 15 (65%). Among 25 patients with adenocarci- patients progressed from high-grade dysplasia to adenocar- noma in Barrett’s esophagus but no visible lesion, the tumor cinoma during endoscopic surveillance. All five patients was limited to the lamina propria (intramucosal) in 22 underwent resection, and although four of the patients had (88%) but invaded into the submucosa in 3 patients (12%). stage 1 tumors, one patient was found to have a stage 3 Of the 25 patients with adenocarcinoma and no visible tumor (T3 N1 M0). Thus, despite a policy of close endo- lesion, 10 underwent en bloc esophagectomy with system- scopic surveillance, invasive cancer developed in five pa- atic thoracic and abdominal lymphadenectomy. From these tients and one had lymph node metastases. Perhaps this risk 10 patients, a total of 370 lymph nodes were removed and is warranted in older patients or those with multiple or analyzed with both routine histology and immunohisto- severe medical comorbidities, but in young fit patients it is chemical staining. Only a single node from one patient was not justified. positive for metastatic cancer. Consequently, we concluded A second approach is to ablate the metaplastic mucosa, that a routine lymph node dissection is unnecessary in including areas of high-grade dysplasia. As discussed pre- patients with high-grade dysplasia or invasive cancer in the viously, techniques that have been used include PDT, elec- absence of an endoscopically visible lesion. However, given trocoagulation, laser, and argon beam coagulation. Long- that in 12% of patients the cancer had invaded deeper than term outcome after ablation by these techniques is not well the mucosa, therapy directed purely at ablation of the mu- defined, and none of these techniques remove a specimen or cosa is not adequate even in the absence of a visible lesion. allow evaluation of the margins of resection. Consequently, In contrast to the rarity of lymph node metastases in an area of invasive cancer might not only go unrecognized, patients with cancer and no visible lesion, we found that but might be inadequately treated as well. Ferguson and 55% of patients with high-grade dysplasia and a biopsy Naunheim96 reviewed three published series involving a showing cancer in the setting of a visible mucosal irregu- total of 46 patients with high-grade dysplasia treated with larity such as an ulcer or nodule had lymph node metasta- PDT. During an average follow-up of less than 2 years, ses.100 Further, in 75% of these patients the tumor had strictures requiring dilation developed in 46% of patients, invaded into or beyond the submucosa. Therefore, the pres- 41% had residual Barrett’s, and 9% had residual high-grade ence of an endoscopically visible abnormality and a biopsy dysplasia. In a series of 100 patients treated with PDT plus of cancer mandates esophagectomy with lymphadenectomy. Nd:YAG laser reported by Overholt et al,87 residual Bar- Obviously, when recommending esophagectomy, one rett’s was present in 57% of patients, and persistent or must balance the potential risks against the benefits of metastatic adenocarcinoma was found in 23% of patients curative resection in patients with high-grade dysplasia. with a pretreatment diagnosis of cancer. Dysplasia of any This requires that the procedure be done at a center that is type persisted in 18% of patients, whereas high-grade dys- experienced with this type of surgery and has a low com- plasia persisted in 10% of patients after treatment. Esoph- plication rate. In addition, it is important that the reconstruc- ageal strictures requiring dilatation developed in 34% of tion in these patients functions well and does not predispose patients. These poor results suggest that PDT should not be them to recurrence of Barrett’s. It is now recognized that considered first-line therapy for Barrett’s and high-grade patients with an intrathoracic anastomosis between their dysplasia, but perhaps could be considered in patients who stomach and residual esophagus are prone to have continued are poor candidates for esophagectomy. reflux of gastric contents into the remaining esophagus. This A third treatment option, and one that we endorse, is has led to esophagitis and even the recurrence of Barrett’s. esophageal resection. Esophagectomy represents the stan- Because this is less likely to occur when the anastomosis is dard against which all other therapies must be compared for done in the neck, we believe that a cervical esophagogastric eradication of cancer. Ferguson and Naunheim96 reviewed anastomosis should be performed routinely. Reconstruction 318 DeMeester and DeMeester Ann. Surg. ● March 2000 with the colon does not lead to this problem, and we believe typically have severe GERD, with significant impairment of that long-term function of a colon graft is superior to that of both LES function and esophageal body motility. Further, a gastric pull-up. the composition of the refluxed gastric juice is different in One technique we now use for high-grade dysplasia is a patients with Barrett’s and complicated Barrett’s esophagus. vagal-sparing esophagectomy with colon interposition. This Patients who reflux both gastric and duodenal juice have a procedure removes the esophagus by the minimally invasive higher prevalence of Barrett’s than do patients who reflux technique of stripping and preserves the intact, innervated gastric juice alone, and among patients with Barrett’s a stomach. The diseased esophagus is replaced with an iso- significantly greater esophageal bilirubin exposure has been peristaltic left colon graft. The operation is very well toler- demonstrated in those with dysplasia. ated, and our early experience suggests that these patients Reflux disease is difficult to control medically in patients have excellent function of both their intact stomach and with Barrett’s, and symptoms are unreliable as a guide. The colonic graft. In addition, they demonstrate few of the side most difficult symptom to control is regurgitation, and there effects potentially associated with a traditional esophagec- is concern that regurgitation and continued reflux of phar- tomy, such as early satiety, dumping syndrome, or diarrhea. macologically altered gastric contents, particularly bile ac- ids in their nonpolar form, may contribute to the develop- ment and progression of Barrett’s. Therefore, it is not SUMMARY surprising that both medical therapy and failed antireflux Since its description in the 1950s the definition of Bar- surgery are both associated with progression of Barrett’s to rett’s esophagus has evolved from the macroscopic visual- dysplasia and adenocarcinoma. In contrast, a functioning ization of gastric-like mucosa in the esophagus to the his- fundoplication appears to provide protection from progres- tologic identification of goblet cells confirming the presence sion of Barrett’s. Intestinal metaplasia extending 3 cm or of intestinal metaplasia within a columnar-lined esophagus. more into the esophagus is unlikely to regress after antire- The length of intestinal metaplasia necessary to be classified flux surgery; however, intestinal metaplasia limited to the as Barrett’s and the relation between intestinal metaplasia of gastroesophageal junction is perhaps more dynamic. We the esophagus and that limited to the cardia are being found that CIM regresses in 73% of patients after fundopli- evaluated. Meanwhile, the definition of Barrett’s has once cation. In addition, we found that low-grade dysplasia re- again become ambiguous. Perhaps the best solution for now gressed in 70% of patients after an antireflux procedure. is to use the term “Barrett’s esophagus” when there is an Antireflux surgery is safe, effective, and durable and often endoscopically visible area of columnar mucosa that on can be performed using minimally invasive techniques. biopsy demonstrates intestinal metaplasia. The term “intes- Thus, antireflux surgery should be strongly considered in tinal metaplasia of the cardia” (CIM) should be used when any patient with Barrett’s. there is a normal-appearing gastroesophageal junction on Patients with Barrett’s and high-grade dysplasia are at endoscopy, but biopsies taken at the gastroesophageal junc- substantial risk for having a coexistent focus of adenocar- tion demonstrate intestinal metaplasia within a columnar cinoma, and even with multiple biopsies there is the possi- epithelium containing mucous cells but devoid of parietal bility of sampling error. Further, adenocarcinoma can de- cells. velop rapidly in the setting of high-grade dysplasia, and if There is evidence that cardiac mucosa is the common the cancer is allowed to invade into the submucosa, 50% of denominator for CIM and Barrett’s and a prerequisite for these patients will have lymphatic metastases, and the pur- intestinal metaplasia. Further, it appears that although car- pose of surveillance will potentially have been negated. diac epithelium is benign, any segment of intestinal meta- Patients with intramucosal adenocarcinoma are unlikely to plasia can undergo dysplastic change and ultimately become have lymphatic metastases; however, current technology a focus of adenocarcinoma. It is logical to expect the degree does not allow precise identification of the depth of wall of risk for developing cancer to be proportional to the penetration superficial to the muscularis propria. Conse- amount of intestinal metaplasia present; however, within a quently, therapy directed only at the mucosa is potentially population, the low risk to any individual is balanced by the inadequate and should not be considered appropriate treat- relative frequency of the process. Thus, given the large ment for cancer. Instead, esophagectomy represents the number of people in America with intestinal metaplasia of standard of care in these patients with a highly curable form the cardia, even a small risk of progression to cancer will of esophageal cancer. result in a large number of patients with adenocarcinoma of Esophagectomy for adenocarcinoma of the esophagus the cardia. This is exactly what is occurring today, with the should be tailored to the patient and the known extent of incidence of adenocarcinoma of the cardia and esophagus disease. For high-grade dysplasia and invasive tumors with- currently rising faster than that of any other cancer in the out an endoscopically visible lesion, a vagal-sparing tech- United States. nique with colon interposition provides an excellent func- The major risk factor for adenocarcinoma of the esoph- tional outcome and removes the entire diseased organ. agus is intestinal metaplasia, and intestinal metaplasia oc- Patients with visible tumors are likely to have invasion curs as a consequence of GERD. Patients with Barrett’s deeper than the mucosa or lymph node metastases. There- Vol. 231 ● No. 3 Columnar Mucosa and Intestinal Metaplasia of the Esophagus 319 fore, those who are physiologically fit should undergo en 21. Meyer W, Vollmar F, Bar W. Barrett-esophagus following total bloc esophagectomy with thoracic and abdominal lymphad- gastrectomy. A contribution to its pathogenesis. Endoscopy 1979; enectomy. In patients with extensive disease or those phys- 11:121–126. 22. Oberg S, Ritter MP, Crookes PF, et al. Gastroesophageal reflux iologically unfit, a palliative transhiatal resection or other disease and mucosal injury with emphasis on short-segment Barrett’s nonsurgical palliative procedure may be appropriate. esophagus and duodenogastroesophageal reflux. J Gastrointest Surg 1998; 2:547–554. 23. Fein M, Ireland AP, Ritter MP, et al. Duodenogastric reflux poten- References tiates the injurious effects of gastroesophageal reflux. J Gastrointest Surg 1997; 1:27–33. 1. Barrett N. Chronic peptic ulcer of the oesophagus and ‘oesophagitis.’ 24. Kuo B, Castell DO. Optimal dosing of omeprazole 40 mg daily: Br J Surg 1950; 38:175–182. effects on gastric and esophageal pH and serum gastrin in healthy 2. Allison P, Johnstone A. The oesophagus lined with gastric mucous controls. Am J Gastroenterol 1996; 91:1532–1538. membrane. Thorax 1953; 8:87–101. 25. Kauer WK, Peters JH, DeMeester TR, Ireland AP, Bremner CG, 3. Barrett M. The lower esophagus lined by columnar epithelium. Sur- Hagen JA. Mixed reflux of gastric and duodenal juices is more gery 1957; 41:881–894. harmful to the esophagus than gastric juice alone. The need for 4. Moersch R, Ellis F, McDonald JR. Pathologic changes occurring in surgical therapy re-emphasized. Ann Surg 1995; 222:525–533. severe reflux esophagitis. Surg Gynecol Obstet 1959; 108:476–484. 26. Chandrasoma P. Pathophysiology of Barrett’s esophagus. Semin 5. Bremner CG, Lynch VP, Ellis HF. Barrett’s esophagus: congenital or Thorac Cardiovasc Surg 1997; 9:270–278. acquired? An experimental study of esophageal mucosal regeneration 27. Lindahl H, Rintala R, Sariola H, Louhimo I. Cervical Barrett’s in the dog. Surgery 1970; 68:209–216. esophagus: a common complication of gastric tube reconstruction. 6. Hayward J. The lower end of the esophagus. Thorax 1961; 16:36–41. J Pediatr Surg 1990; 25:446–448. 7. Bosher L, Taylor F. Heterotopic gastric mucosa in the esophagus with 28. Oberg S, Peters JH, DeMeester TR, et al. Inflammation and special- ulceration and stricture formation. J Thorac Surg 1951; 21:306–312. ized intestinal metaplasia of cardiac mucosa is a manifestation of 8. Paull A, Trier JS, Dalton MD, Camp RC, Loeb P, Goyal RK. The gastroesophageal reflux disease. Ann Surg 1997; 226:522–532. histologic spectrum of Barrett’s esophagus. N Engl J Med 1976; 29. Mason RJ, Oberg S, Bremner CG, et al. Postprandial gastroesopha- 295:476–80. geal reflux in normal volunteers and symptomatic patients. J Gastroi- 9. Haggitt RC, Tryzelaar J, Ellis FH, Colcher H. Adenocarcinoma ntest Surg 1998; 2:342–349. complicating columnar epithelium-lined (Barrett’s) esophagus. Am J 30. Csendes A, Maluenda F, Braghetto I, Csendes P, Henriquez A, Clin Pathol 1978; 70:1–5. Quesada MS. Location of the lower oesophageal sphincter and the 10. Skinner DB, Walther BC, Riddell RH, Schmidt H, Iascone C, De- squamous columnar mucosal junction in 109 healthy controls and 778 Meester TR. Barrett’s esophagus. Comparison of benign and malig- patients with different degrees of endoscopic oesophagitis. Gut 1993; nant cases. Ann Surg 1983; 198:554–65. 34:21–27. 11. Reid BJ, Weinstein WM. Barrett’s esophagus and adenocarcinoma 31. DeMeester TR, Ireland AP. Gastric pathology as an initiator and [review]. Ann Rev Med 1987; 38:477–492. potentiator of gastroesophageal reflux disease [review]. Dis Esoph 12. Clark GW, Ireland AP, Peters JH, Chandrasoma P, DeMeester TR, 1997; 10:1–8. Bremner CG. Short-segment Barrett’s esophagus: a prevalent com- 32. Spechler S, Zeroogian J, Wand H, Antonioli D, Goyal R. The plication of gastroesophageal reflux disease with malignant potential. frequency of specialized intestinal metaplasia at the squamo-colum- J Gastrointest Surg 1997; 1:113–122. nar junction varies with the extent of columnar epithelium lining the 13. Sampliner RE. Practice guidelines on the diagnosis, surveillance, and esophagus. Gastroenterology 1995; 108:A224. therapy of Barrett’s esophagus. The Practice Parameters Committee 33. Mendes de Almeida JC, Chaves P, Pereira AD, Altorki NK. Is of the American College of Gastroenterology. Am J Gastroenterol Barrett’s esophagus the precursor of most adenocarcinomas of the 1998; 93:1028–1032. esophagus and cardia? A biochemical study. Ann Surg 1997; 226: 14. Phillips RW, Wong RK. Barrett’s esophagus. Natural history, inci- 725–735. dence, etiology, and complications [review]. Gastroenterol Clin 34. Griffel L, Weston A, Karalnik N, Amenta P, Das K. Early diagnosis North Am 1991; 20:791–816. of short-segment Barrett’s esophagus (SSBE) using an immunohis- 15. Cameron AJ. Epidemiology of columnar-lined esophagus and ade- tochemical marker: an inter-institutional blinded study. Gastroenter- nocarcinoma [review]. Gastroenterol Clin North Am 1997; 26:487– ology 1998; 114:A140. 494. 35. Fitzgerald RC, Omary MB, Triadafilopoulos G. Dynamic effects of 16. Cameron AJ, Zinsmeister AR, Ballard DJ, Carney JA. Prevalence of acid on Barrett’s esophagus. An ex vivo proliferation and differenti- columnar-lined (Barrett’s) esophagus. Comparison of population- ation model. J Clin Invest 1996; 98:2120–2128. based clinical and autopsy findings. Gastroenterology 1990; 99:918– 36. Oberg S, DeMeester TR, Peters JH, et al. The extent of Barrett’s 922. esophagus depends on the status of the lower esophageal sphincter 17. Eisen GM, Sandler RS, Murray S, Gottfried M. The relationship and the degree of esophageal acid exposure. J Thorac Cardiovasc between gastroesophageal reflux disease and its complications with Surg 1999; 117:572–580. Barrett’s esophagus. Am J Gastroenterol 1997; 92:27–31. 37. Goldblum JR, Vicari JJ, Falk GW, et al. Inflammation and intestinal 18. Clark GW, Ireland AP, DeMeester TR. Dysplasia in Barrett’s metaplasia of the gastric cardia: the role of gastroesophageal reflux esophagus: diagnosis, surveillance and treatment [review]. Dig Dis and H. pylori infection. Gastroenterology 1998; 114:633–639. 1996; 14:213–27. 38. Weston AP, Krmpotich PT, Cherian R, Dixon A, Topalovski M. 19. Stein HJ, Hoeft S, DeMeester TR. Functional foregut abnormalities Prospective evaluation of intestinal metaplasia and dysplasia within in Barrett’s esophagus. J Thorac Cardiovasc Surg 1993; 105:107– the cardia of patients with Barrett’s esophagus. Dig Dis Sci 1997; 111. 42:597–602. 20. Singh P, Taylor RH, Colin-Jones DG. Esophageal motor dysfunction 39. Nandurkar S, Talley NJ, Martin CJ, Ng TH, Adams S. Short-segment and acid exposure in reflux esophagitis are more severe if Barrett’s Barrett’s oesophagus: prevalence, diagnosis and associations. Gut metaplasia is present. Am J Gastroenterol 1994; 89:349–356. 1997; 40:710–715. 320 DeMeester and DeMeester Ann. Surg. ● March 2000

40. Morales TG, Bhattacharyya A, Johnson C, Sampliner RE. Is Barrett’s 60. Peghini PL, Katz PO, Bracy NA, Castell DO. Nocturnal recovery of esophagus associated with intestinal metaplasia of the gastric cardia?. gastric acid secretion with twice-daily dosing of proton pump inhib- Am J Gastroenterol 1997; 92:1818–1822. itors. Am J Gastroenterol 1998; 93:763–767. 41. Pereira AD, Suspiro A, Chaves P, et al. Short segments of Barrett’s 61. Shaffer R, Francis J, Carrougher J, Kadakia S. Effect of omeprazole epithelium and intestinal metaplasia in normal-appearing oesophago- on Barrett’s epithelium after 3 years of therapy. Gastroenterology gastric junctions: the same or two different entities? Gut 1998; 1996; 110:A255. 42:659–662. 62. Ortiz A, Martinez de Haro LF, Parrilla P, et al. Conservative treat- 42. Cooper JE, Spitz L, Wilkins BM. Barrett’s esophagus in children: a ment versus antireflux Surg in Barrett’s oesophagus: long-term re- histologic and histochemical study of 11 cases. J Pediatr Surg 1987; sults of a prospective study. Br J Surg 1996; 83:274–278. 22:191–196. 63. Sampliner RE, Garewal HS, Fennerty MB, Aickin M. Lack of impact 43. Hassall E, Weinstein WM. Partial regression of childhood Barrett’s of therapy on extent of Barrett’s esophagus in 67 patients. Dig Dis Sci esophagus after fundoplication. Am J Gastroenterol 1992; 87:1506– 1990; 35:93–96. 1512. 64. Weston AP, Krmpotich PT, Cherian R, Dixon A, Topalovski M. Prospective long-term endoscopic and histological follow-up of 44. Hoeffel JC, Nihoul-Fekete C, Schmitt M. Esophageal adenocarci- short-segment Barrett’s esophagus: comparison with traditional noma after gastroesophageal reflux in children [published erratum long-segment Barrett’s esophagus. Am J Gastroenterol 1997; 92: appears in J Pediatr 1989 Nov;115(5 Pt 1):836]. J Pediatr 1989; 407–413. 115:259–261. 65. Lagergren J, Bergstrom R, Lindgren A, Nyren O. Symptomatic 45. Qualman SJ, Murray RD, McClung HJ, Lucas J. Intestinal metaplasia gastroesophageal reflux as a risk factor for esophageal adenocarci- is age related in Barrett’s esophagus. Arch Pathol Lab Med 1990; noma. N Engl J Med 1999; 340:825–831. 114:1236–1240. 66. Hameeteman W, Tytgat GN, Houthoff HJ, van den Tweel JG. Bar- 46. Hamilton SR, Yardley JH. Regeneration of cardiac-type mucosa and rett’s esophagus: development of dysplasia and adenocarcinoma. acquisition of Barrett mucosa after esophagogastrostomy. Gastroen- Gastroenterology 1989; 96:1249–1256. terology 1977; 72:669–675. 67. Sharma P, Morales TG, Bhattacharyya A, Garewal HS, Sampliner 47. Cameron AJ, Lomboy CT. Barrett’s esophagus: age, prevalence, and RE. Dysplasia in short-segment Barrett’s esophagus: a prospective extent of columnar epithelium. Gastroenterology 1992; 103:1241– 3-year follow-up. Am J Gastroenterol 1997; 92:2012–2016. 1245. 68. Spechler SJ. Comparison of medical and surgical therapy for com- 48. Provenzale D, Kemp JA, Arora S, Wong JB. A guide for surveillance plicated gastroesophageal reflux disease in veterans. The Department of patients with Barrett’s esophagus. Am J Gastroenterol 1994; of Veterans Affairs Gastroesophageal Reflux Disease Study Group. 89:670–680. N Engl J Med 1992; 326:786–792. 49. Krishnadath KK, Tilanus HW, van Blankenstein M, Bosman FT, 69. Brand DL, Ylvisaker JT, Gelfand M, Pope CE 2d. Regression of Mulder AH. Accumulation of p53 protein in normal, dysplastic, and columnar esophageal (Barrett’s) epithelium after anti-reflux surgery. neoplastic Barrett’s oesophagus. J Pathol 1995; 175:175–180. N Engl J Med 1980; 302:844–848. 50. Menke-Pluymers MB, Mulder AH, Hop WC, van Blankenstein M, 70. DeMeester TR, Attwood SE, Smyrk TC, Therkildsen DH, Hinder Tilanus HW. Dysplasia and aneuploidy as markers of malignant RA. Surgical therapy in Barrett’s esophagus. Ann Surg 1990; 212: degeneration in Barrett’s oesophagus. The Rotterdam Oesophageal 528–542. Tumour Study Group. Gut 1994; 35:1348–1351. 71. Sagar PM, Ackroyd R, Hosie KB, Patterson JE, Stoddard CJ, King- 51. Rice TW, Goldblum JR, Falk GW, Tubbs RR, Kirby TJ, Casey G. snorth AN. Regression and progression of Barrett’s oesophagus after p53 immunoreactivity in Barrett’s metaplasia, dysplasia, and carci- antireflux surgery. Br J Surg 1995; 82:806–810. noma. J Thorac Cardiovasc Surg 1994; 108:1132–1137. 72. Radigan LR, Glover JL, Shipley FE, Shoemaker RE. Barrett esoph- 52. Krishnadath KK, Tilanus HW, van Blankenstein M, et al. Accumu- agus. Arch Surg 1977; 112:486–491. lation of genetic abnormalities during neoplastic progression in Bar- 73. Ransom JM, Patel GK, Clift SA, Womble NE, Read RC. Extended rett’s esophagus. Cancer Res 1995; 55:1971–1976. and limited types of Barrett’s esophagus in the adult. Ann Thorac 53. Cameron AJ, Carpenter HA. Barrett’s esophagus, high-grade dyspla- Surg 1982; 33:19–27. sia, and early adenocarcinoma: a pathological study. Am J Gastro- 74. Starnes VA, Adkins RB, Ballinger JF, Sawyers JL. Barrett’s esoph- enterol 1997; 92:586–591. agus. A surgical entity. Arch Surg 1984; 119:563–567. 75. Williamson WA, Ellis FH Jr, Gibb SP, Shahian DM, Aretz HT. 54. McClave SA, Boyce HW Jr, Gottfried MR. Early diagnosis of co- Effect of antireflux operation on Barrett’s mucosa. Ann Thorac Surg lumnar-lined esophagus: a new endoscopic diagnostic criterion. Gas- 1990; 49:537–542. trointest Endosc 1987; 33:413–416. 76. Attwood SE, Barlow AP, Norris TL, Watson A. Barrett’s 55. Peters JH, Clark GW, Ireland AP, Chandrasoma P, Smyrk TC, oesophagus: effect of antireflux surgery on symptom control and DeMeester TR. Outcome of adenocarcinoma arising in Barrett’s development of complications. Br J Surg 1992; 79:1050–1053. esophagus in endoscopically surveyed and nonsurveyed patients. 77. McDonald ML, Trastek VF, Allen MS, Deschamps C, Pairolero PC, J Thorac Cardiovasc Surg 1994; 108:813–822. Pairolero PC. Barrett’s esophagus: does an antireflux procedure re- 56. van Sandick JW, van Lanschot JJB, Kuiken BW, Tytgat GNJ, Of- duce the need for endoscopic surveillance?. J Thorac Cardiovasc ferhaus GJA, Obertop H. Impact of endoscopic biopsy surveillance of Surg 1996; 111:1135–1140. Barrett’s oesophagus on pathological stage and clinical outcome of 78. DeMeester SR, Campos GM, DeMeester TR, et al. The impact of an Barrett’s carcinoma. Gut 1998; 43:216–222. antireflux procedure on intestinal metaplasia of the cardia. Ann Surg 57. Streitz JM Jr, Ellis FH Jr, Tilden RL, Erickson RV. Endoscopic 1998; 228:547–556. surveillance of Barrett’s esophagus: a cost-effectiveness comparison 79. Low DE, Levine DS, Dail DH, Kozarek RA. Histological and ana- with mammographic surveillance for breast cancer. Am J Gastroen- tomic changes in Barrett’s esophagus after antireflux surgery. Am J terol 1998; 93:911–915. Gastroenterol 1999; 94:80–85. 58. Sampliner RE. Effect of up to 3 years of high-dose lanisoprazole on 80. McCallum R, Polepalle S, Davenport K, Frierson H, Boyd S. Role of Barrett’s esophagus. Am J Gastroenterol 1994; 89:1844–1848. anti-reflux surgery against dysplasia in Barrett’s esophagus. Gastro- 59. Katzka DA, Castell DO. Successful elimination of reflux symptoms enterology 1991; 100:A121. does not insure adequate control of acid reflux in patients with 81. Katz D, Rothstein R, Schned A, Dunn J, Seaver K, Antonioli D. The Barrett’s esophagus. Am J Gastroenterol 1994; 89:989–991. development of dysplasia and adenocarcinoma during endoscopic Vol. 231 ● No. 3 Columnar Mucosa and Intestinal Metaplasia of the Esophagus 321

surveillance of Barrett’s esophagus. Am J Gastroenterol 1998; 93: 92. Ortiz-Hidalgo C, De La Vega G, Aguirre-Garcia J. The histopatho- 536–541. logic and biologic prognostic factors of Barrett’s esophagus: a re- 82. Pearson FG, Cooper JD, Patterson GA, Prakash D. Peptic ulcer in view. J Clin Gastroenterol 1998; 26:324–333. acquired columnar-lined esophagus: results of surgical treatment. 93. van Lieshout EM, Roelofs HM, Dekker S, et al. Polymorphic expres- Ann Thoracic Surg 1987; 43:241–244. sion of the glutathione S-transferase P1 gene and its susceptibility to 83. Naef AP, Savary M, Ozzello L. Columnar-lined lower esophagus: an Barrett’s esophagus and esophageal carcinoma. Cancer Res 1999; acquired lesion with malignant predisposition. Report on 140 cases of 59:586–589. Barrett’s esophagus with 12 adenocarcinomas. J Thorac Cardiovasc 94. Klump B, Hsieh CJ, Holzmann K, Gregor M, Porschen R. Hyper- Surg 1975; 70:826–835. methylation of the CDKN2/p16 promoter during neoplastic progres- 84. Smith R, Hamilton S, Boitnott J, Rogers E. Spectrum of carcinoma sion in Barrett’s esophagus. Gastroenterology 1998; 115:1381–1386. arising in Barrett’s esophagus: a clinicopathologic study of twenty- 95. Wilson KT, Fu S, Ramanujam KS, Meltzer SJ. Increased expression five patients. Lab Invest 1982; 17:78A. of inducible nitric oxide synthase and cyclooxygenase-2 in Barrett’s 85. Levine MS, Caroline D, Thompson JJ, Kressel HY, Laufer I, Her- esophagus and associated adenocarcinomas. Cancer Res 1998; 58: linger H. Adenocarcinoma of the esophagus: relationship to Barrett 2929–2934. mucosa. Radiology 1984; 150:305–309. 96. Ferguson MK, Naunheim KS. Resection for Barrett’s mucosa with 86. Salo JA, Salminen JT, Kiviluoto TA, et al. Treatment of Barrett’s high-grade dysplasia: implications for prophylactic photodynamic esophagus by endoscopic laser ablation and antireflux surgery. Ann therapy. J Thorac Cardiovasc Surg 1997; 114:824–829. Surg 1998; 227:40–44. 97. Messmann H, Endlicher E, Knoechel R, Schoelmerich J. Endoscopic 87. Overholt BF, Panjehpour M, Haydek JM. Photodynamic therapy for fluorescence detection of high- and low-grade dysplasia in Barrett’s Barrett’s esophagus: follow-up in 100 patients. Gastrointest Endosc esophagus after sensitization with 5-amino levulinic acid [abstract]. 1999; 49:1–7. American Gastroenterological Association Poster Session, Digestive 88. Bremner RM, Mason RJ, Bremner CG, et al. Ultrasonic epithelial Diseases Week, Orlando, Florida, May 16–19, 1999. ablation of the lower esophagus without stricture formation. A new 98. Haringsma J, Prawirodirdjo W, Tytgat G. Accuracy of fluorescence technique for Barrett’s ablation. Surg Endosc 1998; 12:342–347. imaging of dysplasia in Barrett’s esophagus [abstract]. American 89. Levine D, Haggitt R, Irvine S, Reid B. Natural history of high- Gastroenterological Association Research Forum, Digestive Diseases grade dysplasia in Barrett’s esophagus. Gastroenterology 1996; Week, Orlando, Florida, May 16–19, 1999. 110:A550. 99. Nigro JJ, Hagen JA, DeMeester TR, et al. Prevalence and location of 90. Schnell T, Sontag S, Chejfec G, et al. High-grade dysplasia (HGD) is nodal metastases in distal esophageal adenocarcinoma confined to the not an indication for surgery in patients with Barrett’s esophagus. wall: implications for therapy. J Thorac Cardiovasc Surg 1999; Gastroenterology 1996; 110:A590. 117:16–25. 91. Williamson WA, Ellis FH Jr, Gibb SP, et al. Barrett’s esophagus. 100. Nigro JJ, Hagen JA, DeMeester TR, et al. Occult esophageal Prevalence and incidence of adenocarcinoma. Arch Intern Med 1991; adenocarcinoma: extent of disease and implications for effective 151:2212–2216. therapy. Ann Surg 1999; 230:433–440.