Barrett Leo R.

Esophagus

Radigan, MD; John

L.

Glover, MD; Frank E. Shipley, MD; Robert E. Shoemaker,

\s=b\ Barrett esophagus is the term describing the presence of an abnormal columnar epithelium in a portion of the esophagus. We have treated 19 patients within the past three years, representing almost 20% of all our esophageal experience. This one pathologic entity has presented as several different clinical pictures: benign stricture, peptic ulceration of the esophagus, intractable esophagitis, and malignancy. One half the patients were under 50 years old, and most were male. The benign lesions have responded well to surgical therapy. There has been an unusually high incidence of malignancy\p=m-\26.3%.The reasons for the infrequent diagnosis of Barrett esophagus are confusion with "short esophagus" and failure to biopsy the proper site.

(Arch Surg 112:486-491, 1977) is lined with columnar "Barrett esophagus," a as rare, curious, and of condition that is questionable significance. During the past three years, however, we have seen 19 patients with this lesion, five of whom had malignancy associated with it. This report summarizes our experience with the Barrett esophagus, giving particular attention to its manner of presentation, the results of treatment, and certain other features that we think are important. of the it is called

portion esophagus Whenepithelium, generally regarded a

a

PRESENTATION Benign Stricture Most of our patients had obstructive dysphagia, and eight of them had benign strictures (Table). The ages of these eight ranged

Accepted

for publication Dec 7, 1976. From the Department of Surgery, Indiana University Medical Center, Wishard Memorial Hospital, Indianapolis. Read before the 84th annual meeting of the Western Surgical Association, Coronado, Calif, Nov 16, 1976. Reprint requests to Wishard Memorial Hospital, Indiana University Medical Center, 1001 W 10th St, Indianapolis, IN 46202 (Dr Glover).

MD

from 35 to 67 years. Five were men, and the duration of their symptoms ranged from three months to 15 years. In all eight patients, the stricture developed at the junction of the normal squamous and the abnormal columnar epithelium (Fig 1), and was at a higher level than usually seen with simple reflux, although all these patients did have reflux, and most had hiatus hernias (Fig 2). On some of the roentgenograms, the reactive stricture formation seemed to have drawn the stomach into the chest (Fig 3 and 4), and we believe that this roentgenographic appearance is the origin of the term "short esophagus," which we believe to be an unfortunate misnomer. In all patients operated on, the esophagus was normal in length and to gross examination. Although we cannot deny the possibility of a short esophagus, we have never seen one. In all patients, the esophageal mucosa appeared normal above the stricture, with biopsy specimens showing squamous epithe¬ lium. Biopsy specimens taken from within or below the stricture, however, showed columnar epithelium, thus confirming the diag¬ nosis.

Peptic Ulcération Two men, ages 47 and 52, had short histories of severe substernal and back pain associated with swallowing. One was anemic and had melena, and roentgenograms in both cases re¬ vealed an eccentric niche in the lower esophagus, as shown in Fig 5. Endoscopy revealed the classical excavating ulcer similar in appearance to a chronic gastric or duodenal ulcer; biopsy revealed columnar epithelium at the margin of the necrotic base of the ulcer. Clinical Presentation of Patients With Barrett Esophagus No. of Patients

Benign stricture Peptic ulcération Severe esophagitis Malignancy Total

Downloaded From: http://archsurg.jamanetwork.com/ by a Western University User on 06/08/2015

5

ÌÌT

Fig 1.—Left and right, Typical benign esophageal stricture at junction of squa¬ mous and columnar mucosa in patient with Barrett esophagus.

2.—Junction of squamous and columnar Barrett esophagus (hematoxylin and eosin,

Fig

Fig 3.—Roentgenographic appearance of "short esophagus" with hiatal hernia and apparent drawing of stomach into chest in patient with Barrett esophagus (left, center, and right).

epithelium 100).

at level of stricture in

patient with

Fig 4.—Roentgenogram of patient shown in Fig 3 after standing and deep breathing. Hernia is reduced, and there is no real shortening of esophagus.

Downloaded From: http://archsurg.jamanetwork.com/ by a Western University User on 06/08/2015

Fig 5.—Typical punched-out Barrett ulcer (arrow) bleeding from peptic ulcération of esophagus.

in

patient with

Incapacitating Esophagitis Three

men

and

one

woman, ages 35 to

55,

were

admitted to the

hospital because of intense heartburn. All had lost weight and had interrupted sleep each night, in spite of drinking up to 1.9 liters of antacid daily. Roentgenographic findings for this group were the least spec¬ tacular in our series, and varied from simple demonstration of reflux to some narrowing of the column of barium in the lower eso¬ phagus. In contrast, the findings at esophagoscopy were astonish¬ ing: in three patients, several centimeters of distal esophagus were thickened, beefy, and friable, and in the fourth it looked necrotic, with a dark surface and little bleeding. A biopsy speci¬ men, however, did show columnar epithelium in this latter patient, as in the others, and cultures of the tissue were negative for fungi and other pathogens in all cases.

Fig 6.—Left and right, Typical roentgenographic findings of carcinoma of esophagus: lesion occurred in area lined by columnar epithelium. RESULTS OF TREATMENT Benign Lesions

Malignancy

Of the 14 patients with benign lesions, all were treated operatively except one patient, who has had a stricture for many years and prefers monthly dilations. The others were treated with fundoplication and with repair of the hernia when it was large. There was no operative mortality and no morbidity, except that three patients required one or two postoperative dilations. Relief of symptoms has remained complete in all patients thus far, although the roentgeno¬ graphic findings do not become completely normal (Fig 7

Five patients had carcinoma, one in the middle and four in the distal one third of the esophagus. All had dysphagia, and roentgen¬ ograms were suggestive of tumor in four of the five (Fig 6). All four patients with lesions in the distal third of the esophagus underwent exploration, and none had tumor in the cardia of the stomach.

Esophagoscopy with biopsy was done six weeks after operation in all 13 operative cases, and has been or will be carried out at yearly intervals since then. In the one patient who had esophagitis that appeared necrotizing, abnormal columnar mucosa had been replaced with squa-

and 8).

Downloaded From: http://archsurg.jamanetwork.com/ by a Western University User on 06/08/2015

Fig 7.—Preoperative roentgenogram in patient with stricture in Barrett esopha¬

Fig 8.—Postoperative roentgenogram in patient shown in Fig 7, after fundoplica¬ tion. Note that appearance of esophagus is not normal, though patient is asympto¬ matic.

Fig 9.—Portion of resected specimen from patient with malignancy. Esophagus has been opened longitudinally, showing color

difference of normal squamous mucosa above and columnar epithelial portion below. Tumor is apparent on right side.

gus.

epithelium by

four months after operation. All the have retained their Barrett esophagus, others, however, and, to date, malignancy has developed in none. mous

Malignant

Lesions

Two of the four patients with malignancy in the lower third had resections with reconstruction by interposition of a jejunal loop. One of the resected specimens is shown in Fig 9. One of the patients who is alive and well 30 months postoperatively had recurrent dysphagia after abdominal repair of hiatus hernia, performed at another hospital. Endoscopie biopsy did not show malignancy, and there was no external evidence of it at thoracoabdominal exploration, except for a slightly bulbous consistency of the strictured area. The other patient is also doing well ten months after

operation. The patient with a midthird lesion underwent a total esophagectomy with reconstruction by a reversed gastric tube two months ago. The other patients had nonresectable lesions, and were treated by radiation; one is alive after 18 months, while the other died as therapy was being initiated.

COMMENT

original article in 1950" was an attempt to then-existing discrepancy between the terms esophagitis and peptic ulcération of the esophagus, and his Barrett's clarify the

clearly showed that whenever a classical, ulcer similar to that seen in the stomach was punched-out in the it always occurred in an area esophagus, present lined by columnar epithelium instead of squamous epithe¬ lium. Although these ulcers behaved like peptic ulcers in the stomach and duodenum, often causing hemorrhage or perforation, their appearance and behavior differed from that of esophagitis due to reflux onto squamous epithelium. Finally, he made the important observation that ectopie islands of gastric mucosa could occur anywhere in the esophagus. Unfortunately, he concluded that this lesion represented a mediastinal stomach, rather than an esopha¬ case

reviews

geal abnormality. A few years later, however, Allison and Johnstonepointed out that the problem was in fact that of an abnormally lined esophagus, and that it was not known whether the condition was congenital or acquired. Subse¬ quent authors have speculated as to its etiology, theorizing (1) congenital rests of embryonic foregut mucosa; (2) ingrowth of gastric-like epithelium after erosion of the squamous lining due to reflux, and (3) proliferation of the submucosal glandular epithelium after reflux has denuded the squamous lining.'-·" Bremner and co-workers presented experimental evidence suggesting that the columnar lining

response to reflux of acid from the stomach,1 and Naef et al," in a large clinical series, believed very strongly that the lesion was acquired and irreversible. We are not

was a

Downloaded From: http://archsurg.jamanetwork.com/ by a Western University User on 06/08/2015

convinced that the latter's evidence is as strong as their opinion, since one of our patients has reverted to a squamous lining. We plan to biopsy all our patients annually to obtain more data and to watch for malignant change, as they advise. One aspect of this condition about which all authors agree is the high incidence of associated malignancy, varying from 8.5% in the series by Naef et al to 26.3% in ours.7"1' Although this represents a large percentage of the esophageal cancer seen in our hospitals during this time, it is difficult to get accurate figures regarding incidence because of the usual assumption that an adenocarcinoma in the esophagus has extended upwards from the stomach. As a matter of fact, statistical surveys by the American Cancer Society automatically consider any adenocarcinoma in the distal 10 cm of esophagus as a cancer of the stomach."1 We believe that our cases and those reported by others clearly refute this theory. Because of the high incidence of malignancy in the Barrett esophagus, a plea for early diagnosis would seem in order, except that the condition is rarely diagnosed at all, much less early. One reason is that a large number of endoscopists fail to realize the necessity of taking biopsy specimens from within and, if possible, beyond an esopha¬ geal stricture, or they assume that any biopsy specimen showing columnar epithelium is from the stomach. Finally, the use of flexible instruments, though a great advance generally, sometimes impedes a thorough evaluation of a

lesion such as this one. Consequently, we urge that all those treating patients with esophageal disease become aware of the serious nature of the Barrett esophagus and insist on complete evaluation and careful follow-up of patients who might have it. That their diagnostic suspicions might be more suffi¬ ciently aroused, we offer the following list of characteris¬ tics of patients with the Barrett esophagus: 1. Severe symptoms of esophagitis that do not respond rapidly and well to standard nonoperative treatment. 2. Complaints of severe substernal or back pain or both associated with swallowing. 3. Acute bleeding of a massive amount from apparent

esophagitis. 4. High esophageal 5. A gus.

strictures.

well-localized, eccentric ulcer in the distal esopha¬

Finally, though our experience is small, we suspect malignant change if a recurrent stricture develops after an apparently adequate operation to correct gastroesophageal reflux. Once the diagnosis has been made, we advise fundo¬ plication, and, if a Barrett ulcer is present, we use a thoracic or thoracoabdominal approach. We intend to continue to observe these patients closely in order to help clarify a confusing pathologic entity that is both more common and more important than previously believed.

References 1. Barrett NR: Chronic peptic ulcer of the oesophagus and oesophagitis. Br J Surg 38:175-182, 1950. 2. Allison PR, Johnstone AS: The oesophagus lined with gastric mucous membrane. Thorax 8:87-101, 1953. 3. Adler RH: The lower esophagus lined by columnar epithelium. J Thorac Cardiovasc Surg 45:13-34, 1963. 4. Bremner CG, Lynch VP, Ellis FH Jr: Barrett's esophagus: Congenital or acquired? An experimental study of esophageal mucosal regeneration in the dog. Surgery 68:209-216, 1970. 5. Burgess JN, Payne WS, Anderson HA, et al: Barrett esophagus: The columnar-epithelial-lined lower esophagus. Mayo Clin Proc 46:728-734, 1971.

6. Naef AP, Savary M, Ozzello L: Columnar-lined lower esophagus: An lesion with malignant predisposition. J Thorac Cardiovasc Surg 70:826-833, 1975. 7. Hawe A, Payne WS, Weiland LH, et al: Adenocarcinoma in the columnar epithelial lined lower (Barrett) oesophagus. Thorax 28:511-514, 1973. 8. Shafer RB: Adenocarcinoma in Barrett's columnar-lined esophagus. Arch Surg 103:411-413, 1971. 9. Stemmer EA, Adams WE: The incidence of carcinoma at the esophagogastric junction in short esophagus. Arch Surg 81:771-780, 1960. 10. Clinical Staging System for Carcinoma of the Esophagus. American Cancer Society. Cancer J Clin 25:50, 1975.

acquired

Discussion

Raymond C. Read, MD, Little Rock, Ark: We have also been interested in this unusual lesion. An understanding of it is helpful in difficult diagnostic problems in the lower esophagus and the upper stomach. A patient of ours had the type of lesion that is the most common presention of the Barrett esophagus, that is, a benign lesion, reflux esophagitis, with a peptic stricture of the esophagus. The stricture was laid open. Adenomatous tissue lined the lower esophagus in the region of the stricture, interposed with ulcération and now and then islands of squamous epithelium.

I think that most of us would agree, as Dr Radigan said, that this is an acquired lesion, and that this tissue is metaplastic. As Dr Ellis showed experimentally, in the lower esophagus, if the squa¬ mous epithelium is digested with acid, there is a tendency for adenomatous metaplasia to occur. The microscopy in our patient was very disorganized, with no real pattern. It tended to be hemorrhagic, and it was a lesion consistent with metaplasia.

However, there are other cases of Barrett esophagus, not as common, that I believe are columnar epithelium lining on a con¬ genital basis. We know that all of us, through embryonic life, pass

through a phase where the esophagus is lined with columnar epithelium, and only later does it cavitate and become squamous. We found a high stricture in a middle-aged man, and all of the distal two thirds of esophagus, from the aortic arch downward, was lined with adenomatous epithelium. There was a small hiatus hernia, but the important thing was that there was no inflamma¬ tion in this area at all. There was some inflammation high at the junction of the squamous and adenomatous epithelium. During left thoracotomy, we saw that the epithelium lining the distal esophagus was all adenomatous and salmon pink in color. It was not hemorrhagic. There was no esophagitis in this area, and there is no esophagitis distally. The biopsy specimen from this man showed a very orderly epithelium with glands. There was no hemorrhage.

Downloaded From: http://archsurg.jamanetwork.com/ by a Western University User on 06/08/2015

some people are born with an adenomatous lining esophagus, and this does not come to our attention unless stricture develops or they occasionally bleed. We have oper¬ ated on one or two patients for massive upper gastrointestinal hemorrhage who have shown at operation the beefy adenomatous epithelium, as Dr Radigan described. The beefy red lesion with

I believe that

to their

of Barrett ulcers. All of us who take care of patients with gastroesophageal lesions of the upper stomach and lower esophagus should be aware of it. Lawrence DenBesten, MD, Iowa City: In their presentation, Dr Radigan and his colleagues have placed the Barrett esophagus in proper perspective. Their data suggest that (1) its incidence may be far greater than has been appreciated from studies using less intense diagnostic efforts; (2) the potential of this lesion for malignant degeneration recommends vigorous diagnostic and therapeutic attention to the patient with odynophagia; (3) the confusing and controversial literature on the etiology of this lesion may be clarified by the study of appropriate animal models and frequent biopsy of human disease during injury and repair; and (4) the presence of columnar epithelium in the esophagus is univer¬ sally associated with and may be caused by gastroesophageal reflux, a condition for which surgeons now have an effective array of operations. The authors' hypothesis that the Barrett esophagus represents an injury and repair phenomenon is reasonable. They have ob¬ served a return to squamous epithelial lining after correction of reflux. Epithelium in the tracheobronchial tree subjected to long¬ standing injury, as in cigarette smoking, is also replaced by columnar epithelium, but may heal when cigarette smoking is abandoned. Experiments in our laboratories have rather clearly shown that both the animal and human esophagus is injured by low concentrations of hydrogen ion accompanied by bile salts, or high concentrations of hydrogen alone, and that chronic injury results in columnar metaplasia. Certain questions of interest are not addressed in the article: (1) Do the authors have data on manometry, esophageal pH, and the Bernstein test in the preoperative and postoperative period? (2) Have the authors found the various operations to correct gastro¬ esophageal reflux equally effective? (3) Why have the strictures responded so well to one or two dilations when most surgeons find they must continue to dilate or calibrate the strictured esophagus for at least six to eight weeks after fundoplication for classic reflux esophagitis? Finally, careful endoscopie examination with biopsy at and below the region of stricture or ulcer in patients with dysphagia is crucial for the diagnosis of this lesion. Because the influence of antireflux operations on the malignant potential of the Barrett esophagus is unknown, repeat examinations should be continued as long as columnar epithelium persists. erosions

are cases

C. G. Wheeler, MD, Dallas: We are seeing more of this entity that we are employing esophagoscopy routinely in the work up of patients with lower esophageal problems. We have had to deal with a number of patients who have had prior hiatal hernia operations using the Belsey and the Hill repairs and who still had now

reflux problems. We have a different technique than the usual for repairing this, and I would like to describe it. We want to increase the length of the esophagus and lower the cardia, and so we form a gastric tube. Instead of a reverse gastric tube, it is a direct gastric tube along the lesser curvature. This defuses or inactivates many of the pressure effects in the intragastric and extragastric areas against the cardia, and it is a repair that can be done without any tension whatsoever. A group of patients treated in this way now have been followed up for periods of up to four years, and they are able to eat normally and to vomit if necessary. They are rid of their reflux symptoms, and postoperative biopsies have shown reversion back to normal squamous epithelium. In the repair, we close the crura in the Allison manner, and then use a suture line, suturing the anterior and posterior walls of the stomach alongside a #50 Maloney bougie to form a gastric tube along the lesser curvature, lowering the cardia to the level of the incisura. Dr Glover: Dr Read, we agree with what you say, and we believe also that some of these cases may be congenital in origin. That has been suggested by some of the earlier work, I think both by Mr Barrett and Mr Allison, and we think that may be the

severe

case.

Dr DenBesten, we agree with your comments also. We have not done manometric studies, but some were reported on by the Mayo Clinic, indicating that that portion of the esophagus above the stomach had normal motility. We have not done the Bernstein test, and our experience is only with the Nissen type fundoplication. I have no idea why some patients only required one or two post¬ operative dilations. They have undergone operation with a sizable dilator in place, and one patient, whose roentgenogram shows narrowing, swallows so well that he won't accept further treat¬ ment.

Dr Wheeler, we appreciate your comments also, and find your technique interesting. Finally, I would like to say that it does require a lot of patience and a lot of time to try to obtain the biopsy specimens from some of these patients with severe strictures. That sort of patience and time has been given in this study by Dr Radigan, who really de¬ serves the credit for the majority of the work. One thing that is underscored by the comments of all the discussants is that sur¬ geons shouldn't leave endoscopy to other people by default. It is still an important surgical technique.

Downloaded From: http://archsurg.jamanetwork.com/ by a Western University User on 06/08/2015

Barrett esophagus.

Barrett Leo R. Esophagus Radigan, MD; John L. Glover, MD; Frank E. Shipley, MD; Robert E. Shoemaker, \s=b\ Barrett esophagus is the term describi...
7MB Sizes 0 Downloads 0 Views