What is the evolution of Barrett's mucosa after antireflux surgery?
H. Hölscher, C. Gutschow, K.T.E. Beckurts, P. Schneider (Cologne)
In a retrospective study concerning the change of Barrett's mucosa after antireflux surgery, patients with the following criteria were included: gastroesophageal reflux disease with Barrett's esophagus (BE) and antireflux surgery comprising Nissen fundoplication, Angelchik prosthesis or Roux-en-Y duodenal diversion. BE was defined endoscopically, if the columnar epithelium between the anatomical cardia and the squamocolumnar border had at least an extent of 3 cm including tongue-like lesions according to type II of Savary [1, 2].
From 1982 to 1993, 65 patients have been included in the study. In all cases gastroesophageal reflux disease had been proven preoperatively by patient history, endoscopy, esophageal 24-hour-pH-monitoring, and esophageal manometry. These were 49 males and 16 females with an average age of 53.6 (range 24-82) years. 51 of these patiens had had primary antireflux operations (fundoplication n = 50, Angelchik prosthesis n = 1), whereas 14 had had reoperations after failed fundoplication (refundoplication n = 13, Roux-en-Y gastric resection n = 1).
The degree of reflux esophagitis prior to antireflux surgery showed in addition to Barrett's mucosa in 30 cases (46.2%) degree IV consisting of 26 stenosis and 4 Barrett ulcers. In 8 cases (12.3%) a degree III esophagitis and in 27 cases (41.5%) a degree II esophagitis was found [1, 3]. The extent of BE was measured endoscopically as the distance from the most orad part of the squamocolumnar border to the aborad end of the tubular esophagus. The average length of BE prior to antireflux surgery (n = 63 patients) rised to 5.75 cm (SEM ± 0.39, range 3-15 cm). In most cases the specialized epithelium of the columnar cell lined lower esophagus was proven by histological examination of biopsies.
The clinical course of the patients was prospectively documented after antireflux surgery in a follow-up program. The patients were asked for yearly examinations including registration of symptoms, upper gastrointestinal endoscopy, esophageal manometry and 24-hours-pH-monitoring. The extent of BE after antireflux operation was measured endoscopically during the reexamination as mentioned above. Not all patients did participate in this yearly follow-up program or to all diagnostic procedures. However, 58 of 65 patients (82.2%) were taking part in the follow-up program and had at least
1 reexamination. 7 patients who did not come to the reexaminations were contacted by phone or via the family doctors. Altogether, concerning the question of development of cancer in Barrett's esophagus after antireflux surgery, the course of 100% of the patients is known with a reendoscopy rate (>= 1 year after surgery) up to 1995 of 89% (n = 58) of the patients. For 2 patients no primary measurement of BE was available. For 8 patients, the extent of BE had not been exactly documented during reendoscopy. That means that, for the analysis of regression of BE, information on 48 patients was available. A change in the extent of BE was defined as an increase or decrease of at least 1 cm.
The time of follow-up between the first antireflux operation and the end of the study or time of detection of a carcinoma averaged to 8.1 years (range 2.4-25 years) for the 65 Barrett's patients. A postoperative recurrence of reflux was documented in 6 cases (9.2%). In 3 cases this was due to a telescope phenomenon, in 2 further cases due to loosening of the cuff, and in 1 case due to a perforation of an Angelchik prosthesis into the stomach. 5 of these patients had a successful refundoplication, 1 had medication with omeprazole. Statistics comprised the calculation of mean and standard error of the mean (± SEM) as the measuring values showed a normal distribution. For analysis of significance, the 2-tailed T-test for paired samples was applied.
In 48 of the patients at least one further exact measurement of the extent of the columnar epithelium of the distal esophagus was available on the average 4.2 years (range 1-11.5) after the initial evaluation. At the time of the last measurement the average length of Barrett's esophagus of these 48 patients was 5.50 cm (SEM ± 0.4, range 3-15 cm) compared to 5.62 cm (SEM ± 0.41, range 3-15 cm) at the initial measurement. This represents no significant difference (p = 0.55). In only 9 of these patients a change (>= 1 cm) of the extent of the esophagus was found, whereas in 39 the length of BE was reported to be the same (± < 1cm) as in the first endoscopy. The alteration in these patients was 6 x an increase of 1-2 cm and 3 x a decrease of 1-5 cm.
During the follow-up period 2 patients of the series of 65 developed a carcinoma after the antireflux operation. Both patients had been documented thoroughly in their course with endoscopy, manometry, and 24-hour-pH-monitoring prior and after fundoplication as well as prior to esophagectomy (Table I).
The evaluation of the literature shows that 21 authors have observed a regression of BE after antireflux surgery, whereas 7 authors did not (Tables II and III). Most of the regressions, however, were only partial and they were observed only in a small percentage of patients of these series. Only in 32 mostly not very well documented cases a complete regression has been reported. The best documented reports by Brand and Skinner show that after successful antireflux surgery a regression of the columnar epithelium is possible [14, 16]. In both studies the patients in which a regression of BE was observed, had no or only physiologic postoperative gastroesophageal reflux in 24-hour pH-monitoring. Brand has concluded that a regression of the columnar epithelium is correlating with the effect of antireflux surgery. His results, however, have been severely criticized because the histologically examined biopsies of columnar epithelium had been taken by blind suction biopsy . The study of Skinner shows that in spite of successful antireflux operation, the regression of BE is not mandatory. Only 2 of his 13 examined patients showed a regression, although a good result of the operation had also been confirmed in the other 11 patients by 24-hour pH-monitoring. It is interesting that in 2 further patients of this group dysplasias of BE which had been found preoperatively, showed a reduction after successful antireflux operation. McCallum has reported the same effect .
If the results of the mentioned studies (Tables II and III) are evaluated in total, a mostly partial regression was observed in 89 of 512 patients (17.4%: 11.1% partial, 6.3% complete). Again, only 32 of these regressions, i.e. 36%, were complete.
A survey of the literature revealed 74 cases of adenocarcinoma which developed in BE after different types of antireflux surgery. These case-reports are mostly mentioned within series on surgical results of esophageal carcinoma or results of antireflux surgery. They are, therefore, not documented in detail. One important criterion for an evaluation of these cases is the efficiency of antireflux surgery as patients with a failure of this procedure have to be considered as not different from patients with BE without antireflux surgery.
Another problem is the interval between antireflux operation and detection of carcinoma in BE which ranges from 1-15 years. If this period is too short, e.g. 1-3 years, an early cancer may have been missed prior to antireflux surgery and later may have developed into clinical evident cancer.
The extent of BE mostly remains unchanged after successful antireflux surgery. In approximately 11% of the cases, partial regression has been observed whereas a complete regression is rare (< 6%). The development of an adenocarcinoma in BE is still possible after successful antireflux operation. The risk for malignant degeneration in spite of surgical elimination of gastroesophageal reflux seems not to be reduced compared to patients without antireflux operation.
2. Savary M, Ollyo JB, Monnier P. Frequency and importance of endobrachyesophagus in reflux disease. In: Siewert JR, Hölscher AH, eds. Diseases of the esophagus. Berlin, Heidelberg, New York: Springer, 1988:529-536.
18. Pollara WM, Zilberstein B, Cecconello I, Vencon FE, Parada AA, Pinotti HW. Barrett's esophagus. Late results of conservative management. In: Siewert JR, Hölscher AH, eds. Diseases of the esophagus. Berlin, Heidelberg, New York, Rome, Tokyo: Springer, 1988:545-549.
19. Conti Nibali S, Barresi G, Tuccari G, Rivosecchi M, Magazzù G. Barrett's esophagus in an infant: a long standing history with final postsurgical regression. J Ped Gastroenterol Nutr 1988;7:602-607.
21. Henrion J, Shapira M, Pourbaix A, Heller FR. Régression complète d'un endobrachy-sophage (sophage de Barrett) après correction du reflux biliaire chez un malade gastrectomisé. Gastroenterol Clin Biol 1989;13:745-746.
23. Ollyo JB, Gonvers JJ, Froehlich F, Restellini A, Monnier P, Fontolliet C, Savary M. L'endobrachy-sophage régresse-t-il après traitement efficace du reflux gastro-sophagien? Schweiz Med Wschr 1990;120:716-720.
29. Salo JA, Nemlander A, Färkilla M, Kiviluoto T, Kivilaakso E. Treatment of Barrett's metaplasia by antireflux surgery and endoscopic laser ablation: clinical experience in 10 patients. Gastroenterology 1995;108:A1243.
30. Ortiz A, Martinez de Haro LF, Parrilla P, Morales G, Molina J, Bermejo J, Liron R, Aguilar J. Conservative treatment versus antireflux surgery in Barrett's esophagus: long-term results of a prospective study. Br J Surg 1996;83:274-278.