Primary Motility  Disorders of the  Esophagus
 The Esophageal
 Mucosa
 The
 Esophagogastric  Junction
 Barrett's
 Esophagus

  Browse by Author
  Browse by Movies
OESO©2015
 
Volume: Barrett's Esophagus
Chapter: Etiology and origins of Barrett's epithelium
 

How abnormal is gastric function in patients with Barrett's esophagus?

C. Scarpignato, I. Pelosini, S. Contini (Parma)

Barrett's esophagus is a complication that develops in a minority of patients with gastroesophageal reflux disease (GERD). It would be of little interest, were it not for its well-established association with adenocarcinoma of the esophagus, a cancer whose incidence has increased rapidly since the 1970s and remains on the rise [1]. Although our understanding of Barrett's esophagus has improved considerably in recent years, much remains unknown.

Barrett's esophagus is an acquired condition resulting from severe esophageal mucosal injury. However, it still remains unclear why some patients with GERD develop Barrett's esophagus whereas others do not. Animal studies show that excision of esophageal mucosa in conjunction with reflux of either acid alone or acid and bile results in re-epithelialization of the esophagus with primarily columnar epithelium, whereas excision of the esophageal mucosa alone results in re-epithelialization primarily with squamous epithelium [2, 3]. Bile reflux alone does not cause columnar re-epithelialization in animal studies [3].

In humans, Barrett's esophagus is clearly associated with severe gastroesophageal reflux (GER). Compared with patients with erosive and nonerosive GERD without Barrett's esophagus, patients with columnar-lined esophagus (CLE) typically have greater esophageal acid exposure based on 24-hour pH monitoring [4-6]. However, there is some overlap and some studies found no difference in intraesophageal acid exposure time between Barrett's patients and patients with high-grade esophagitis [7]. Part of the increase in acid exposure in Barrett's patients may be related to the almost uniform presence of a hiatal hernia, which is typically longer and associated with larger defects in the hiatus than controls or patients with esophagitis alone [8]. In addition, patients with Barrett's esophagus have a lower basal esophageal sphincter pressure (LESP) compared with GERD patients without columnar metaplasia [5].

Gastric volume is the major determinant of the amount of fluid that can be refluxed: an increased gastric volume increases the likelihood of GER. The volume of intragastric contents reflects the balance between inputs and outputs. The inputs results from swallowing of foods and saliva, gastric secretion, and duodenogastric reflux while the outputs correspond to GER and gastric emptying (Figure 1). Under basal fasting conditions, normal subjects have a steady state resting volume of approximately 10 ml, which results from secretory and emptying rates of 1 ml/min. After a meal, intragastric volume increases by the amount of food and liquids swallowed, which can be 1,000 ml or more. During the early postprandial period, the contribution of gastric output is negligible. While emptying of the meal progresses, however, the participation of fluid output becomes more and more important and eventually represents the total intragastric volume (i.e. 10 ml) when the fasting steady state level is reached, usually in 3 to 6 hours [9].

It is therefore evident that the increased esophageal exposure to acid in patients with Barrett's esophagus could also be the consequence of gastric hypersecretion and/or delayed gastric emptying, which both increase the volume and the aggressive power of the refluxate. The presence of increased acid secretion in this patient population may also

Figure 1. Schematic representation of fluid dynamics within the upper gastrointestinal tract (from [9]).

explain why these patients do not respond adequately to standard doses of H2-receptor antagonists [10, 11] or proton pump inhibitors (PPIs) [11-13]. Gastric secretion and emptying are two interrelated functions that cannot be considered or studied independently from each other [14]. From a physiological point of view, indeed, it is well known that the main stimulants of acid secretion (namely acetylcholine, histamine and gastrin) all modify gastric emptying by a mechanism that may be independent of their secretory effect [15]. Similarly, antisecretory drugs can also affect gastric emptying rate [15]. The aim of this paper is to present the available data on gastric secretion and emptying in patients with Barrett's esophagus and discuss their relevance to the pathogenesis of the disease.

Gastric secretion

The presence of basal acid hypersecretion in a small group of patients with Barrett's esophagus was first reported by Mulholland et al. [16]. Although in this patient population meal-stimulated secretion was comparable to that of healthy controls, gastrin (G-17-I)induced acid hypersecretion was significantly higher when compared with controls. However, parietal cell sensitivity to gastrin, expressed as the concentration causing halfmaximal acid secretion, was not significantly different (i.e. 56.8 ± 5.5 fmol/ml versus 63.6 ± 6.9 fmol/ml, respectively) [16]. In this study no data were provided on how many patients were true hypersecretors (i.e. with basal acid output - BAO - higher than 10 mEq/h) and how many had concomitant esophagitis and/or duodenal ulcer. In a subsequent larger study, Collen et al. [10] reported that patients with Barrett's esophagus display - as a group - a significantly higher BAO than healthy controls. Many patients however had their value in the normal range (Figure 2). Only 15 out of 42 patients with Barrett's esophagus (i.e. 36%) had BAO higher than 10 mEq/h. There was a significant correlation between BAO and the daily ranitidine dose required for esophageal mucosal healing and complete symptom control (Figure 2). Although in a previous study of the same group of investigators [17] no difference between patients with Barrett's esophagus and those with reflux esophagitis (without columnar metaplasia) in either BAO mean values and proportion of subjects with BAO higher than 10 mEq/h was found, Stein et al. [18] reported that basal and pentagastrin-stimulated acid secretion were in Barrett's patients significantly higher than those found in symptomatic patients with erosive esophagitis. Different selection criteria of the two patient populations may explain the different findings of the two investigations. In studying acid secretion, there are indeed many confounding influences, such as sex (male secrete more than females) and the presence or esophagitis, duodenal ulcer or Zollinger-Ellison syndrome (ZES) [19]. In a landmark study which adopted appropriate controls (i.e. patients with esophagitis, but not CLE, who were matched for age, sex and presence or absence of duodenal ulcer or ZES), Hirschowitz [20] did unquestionably show that there is no significant difference in both basal and stimulated secretion between patients with esophagitis and those who have in addition Barrett's esophagus (Figure 3). The presence or absence of complications (e.g. stricture, ulceration, dysplasia or cancer) does seem to influence neither acid secretion nor the proportion of Barrett patients with

Figure 2. Correlation between basal acid output (BAO) anddaily raniditine dose required for healing of esophagitis in patients with Barrett's eso-phagus (from [10]).

Figure 3. Gastric acid and fasting serum gastrin in patients with Barrett's esophagus and appropriately matched controls (from [20]). DU: duodenal ulcer; BAO: basal acid output; PAO: peakacid output

hypersecretion [21]. The available results from gastric secretory studies, including the proportion of patients with basal acid hypersecretion, are summarized in Table I.

Table I. Gastric acid secretion in Barrett's esophagus: summary of the published studies

The role of pepsin has not been explored extensively in Barrett's esophagus. Activation of pepsin requires a pH of less than 3 [22]; thus, separating the role of pepsin from acid alone is difficult. One early study [23] reported an increased gastric juice concentration of pepsin, but it is flowed by methodological drawbacks. And, indeed, Hirschowitz [20] could not find any difference in pepsin output between patients with Barrett's esophagus and appropriate controls.

The gastric pH profile in this patient population was first studied in 1985 by Stein et al.[18] who found that - compared to patients with esophagitis - those with columnar metaplasia displayed a marked shift of the recorded pH values towards a more acidic range. The development of better computer programs to quantitate and analyze 24-hour pH recordings has allowed Savarino et al. [24] to study thoroughly the intragastric acidity pattern of patients with Barrett's esophagus. They found no difference in the 24-hour and daytime patterns of gastric pH between patients and healthy subjects, a finding recently confirmed by Fein et al. [25]. However, during the night, acidity was significantly lower in Barrett's patients (Figure 4).

It must be emphasized that almost all the above secretory studies have been performed before the "Helicobacter revolution". And the more recent pH-metric investigations did not specifically look at the Helicobacter status, which does markedly influence the acid secretory capacity of the stomach [26, 27] as well as the effect of antisecretory drugs,

Figure 4. 24-hours mean gastric pH profiles in patients with Barrett's esophagus and healthy controls (from [24]).

mainly PPIs [28, 29]. Furthermore, although still debated, H. pylori infection may protect against the development of esophagitis by decreasing the potency of the gastric refluxate in patients with corpus-predominant gastritis [30, 31]. It is therefore evident that further studies taking into account the Helicobacter status are needed to definitely establish whether patients with Barrett's esophagus have a gastric secretory pattern different from that seen in patients with GERD or healthy controls.

Gastric emptying

On theoretical grounds, a delayed emptying rate can be followed by an increased number and duration of GER episodes. As a consequence of the increased intragastric volume, there will be indeed an increase of gastroesophageal pressure gradient, an increase of the number of spontaneous lower esophageal sphincter (LES) relaxations and, finally, a stimulation of acid secretion with subsequent increase of the gastric content available for reflux. Furthermore, delayed clearance of bile acids and pancreatic secretions, which pass intermittently across the pylorus, may permit these toxic agents to reflux into the esophagus and to cause esophageal mucosal damage [32].

It is well known to every clinician that some patients with GERD complain of vomiting [33, 34], a symptom suggesting subclinical or overt gastric retention [35]. More than 25 years ago DeMeester et al. [36] were among the first to suggest that symptomatic postprandial GER may originate from gastric stasis. Studies concerning gastric emptying in GERD (for review see [32]), however, have provided conflicting results. Many reasons can account for discrepancies in findings [37, 38], the most striking one being a different methodological (i.e. different techniques of measuring gastric emptying and different kinds of test meal) and statistical (different methods of testing for significance between groups in gastric emptying studies) approach. In addition, criteria for selection of patient and control populations were also different in the different investigations. In particular, patients had a different gravity of the disease (with or without esophagitis), the range of control values was often wide, controls were not often age and sex matched and the influence of associated diseases (especially dyspepsia) [37, 39] was never taken into account. Patients with reflux symptoms are indeed not a homogeneous population and other digestive disorders can mimic GERD [40]. Last but not least, when doing this kind of investigation one should keep in mind that motor derangement can be intermittent.

Although vomiting and other dyspeptic symptoms are suggestive of gastric stasis, their predictive value for the presence of documented gastric stasis is very poor [41]. Scintigraphic studies (or other validated methods for measurement of gastric emptying) are therefore the only reliable way of evaluating the frequency of gastric stasis in GERD [42]. Of 20 studies including 670 patients, in which objective evaluation of gastric emptying was performed, 14 concluded that a significant delay in gastric emptying was present in GERD patients. Amongst 10 investigations, which studied simultaneously gastric emptying of both solids and liquids, three found no delay in gastric emptying, four that stasis was present for liquids and solids, two that stasis was present for solids only and one that stasis occurred with liquids only [43]. The frequency of delayed gastric emptying in patients with GERD, available from 16 studies, is given in Table II. Gastric stasis was found in about 40% of patients and affected solids a little more than liquids. Conversely from data obtained in GERD, the results from studies performed in patients with Barrett's esophagus [16, 18,

Table II. Frequency of gastric stasis (expressed as % of patients with delayed gastric emptying) in GERD (from [41]).

44, 45] uniformly show that the emptying rate of both solids and liquids fall within the normal range (Table III), a finding which is not affected by the presence of complications [21]. In the only scintigraphic investigation [44] reporting the distribution of emptying half-times in the patient population studied, 12 out of 17 (i.e. 70%) Barrett patients had a normal gastricemptying of solids while only 2 of them displayed a delayed emptying and 3 actually showed an acceleration of the emptying rate (Figure 5).

Table III. Gastric emptying in Barrett's esophagus. Summary of the published studies.

Recent evidence suggests that - compared to healthy subjects - in GERD patients there is a more pronounced post-prandial fundic relaxation [46, 47] and that it is the emptying of both solids and liquids from the proximal stomach that is specifically delayed [47] in this clinical condition. Slow proximal but not slow distal or total gastric emptying correlated with increased 24-hour and post-prandial esophageal acid exposure [48]. Taking these data into account, it would be worthwhile to study in Barrett patients either the emptying of proximal and distal stomach as well as intragastric distribution of food to better investigate the gastric motor pattern in this particular patient population.

Figure 5. Gastric emptying of solids in symptomatic patients with Barrett's esophagus (from [45]).

Conclusion

Although not a constant finding, gastric acid hypersecretion is relatively common in patients with Barrett's esophagus, irrespective of the presence of esophagitis. This finding, however, could be simply attributed to the frequent but coincidental presence of duodenal ulcer [20]. Indeed, studies, which adopted appropriate controls, failed to show any difference in gastric secretory states.

Available data are consistent with the idea that a defective gastroesophageal barrier is the predominant factor for the development of complications in both GERD and Barrett's esophagus [49]. Indeed, in a population of 75 GERD patients (including 13 patients with CLE), gastric hypersecretion was a less frequent finding than a mechanically defective LES [50]. In the presence of normal gastric secretion, complications occurred in 18% of those with a normal sphincter and 77% of those with a defective sphincter while, in the presence of hypersecretion, the complication rate was 40% and 82%, respectively [50]. In addition, although abnormal esophageal exposure to acid (which correlates to abnormal sphincter pressure [51]) was detected in almost all (i.e. 93%) Barrett patients, gastric hypersecretion was present only in 44% of them [21]. All these findings suggest that the development of reflux complications (including Barrett's esophagus) is mainly dependent on the presence of a defective barrier rather than of acid hypersecretion. Finally, the low prevalence of columnar metaplasia in patients with ZES [52], whose acid secretion is more than 8 folds that of healthy subjects, emphasizes once again the limited role of gastric hypersecretion alone in the pathogenesis of Barrett's esophagus. Nonetheless, inhibition of acid secretion still remains a cornerstone of the therapy due to its ability to reduce the aggressive power of the refluxate [53, 54].

Like gastric secretion, gastric emptying is almost always normal in patients with Barrett's esophagus, although further studies examining the behavior of the proximal and distal stomach are needed. These data, and the little effect that prokinetics have on defective esophageal peristalsis and sphincter pressure and function (e.g. TLESRs) [55], raise some doubt on the use of these agents in Barrett's disease.

References

1. Spechler SJ. Barrett's esophagus. New Engl J Med 2002;346:836-842.

2. Bremner CG, Lynch VP, Ellis FH. Barrett's esophagus: congenital or acquired? An experimental study of esophageal mucosal regeneration in the dog. Surgery 1970;68:209-216.

3. Gillen P, Keeling P, Byrne PJ, West AB, Hennessy TP. Experimental columnar metaplasia in the canine oesophagus. Br J Surg 1988;75:113-115.

4. Neumann CS, Cooper BT. 24 hour ambulatory oesophageal pH monitoring in uncomplicated Barrett's oesophagus. Gut 1994;35:1352-1355.

5. Singh P, Taylor RH, Colin-Jones DG. Esophageal motor dysfunction and acid exposure in reflux esophagitis are more severe if Barrett's metaplasia is present. Am J Gastroenterol 1994;89:349-356.

6. Oberg S, DeMeester TR, Peters JH, Hagen JA, Nigro JJ, DeMeester SR, Theisen J, Campos GM, Crookes PF. The extent of Barrett's esophagus depends on the status of the lower esophageal sphincter and the degree of esophageal acid exposure. J Thorac Cardiovasc Surg 1999;117:572-580.

7. Coenraad M, Masclee AA, Straathof JW, Ganesh S, Griffioen G, Lamers CB. Is Barrett's esophagus characterized by more pronounced acid reflux than severe esophagitis? Am J Gastroenterol 1998;93:1068-1072.

8. Cameron AJ. Barrett's esophagus: prevalence and size of hiatal hernia. Am J Gastroenterol 1999;94:2054-2059.

9. Dubois A. Role of gastric factors in the pathogenesis of gastroesophageal reflux: emptying, acid and pepsin, bile reflux. In: Castell DO, Wu WW, Ott DJ, eds. Gastroesophageal reflux disease. pathogenesis, diagnosis, therapy. New York : Futura Publishing Co, 1985:81-97.

10. Collen MJ, Johnson DA. Correlation between basal acid output and daily ranitidine dose required for therapy in Barrett's esophagus. Dig Dis Sci 1992;37:570-576.

11. Srinivasan R, Katz PO, Ramakrishnan A, Katzka DA, Vela MF, Castell DO. Maximal acid reflux control for Barrett's oesophagus: feasible and effective. Aliment Pharmacol Ther 2001;15:519-524.

12. Katzka DA, Castell DO. Successful elimination of reflux symptoms does not insure adequate control of acid reflux in patients with Barrett's esophagus. Am J Gastroenterol 1994;89:989-991.

13. Savarino V, Mela GS, Zentilin P, Celle G, Vigneri S. Once or twice daily doses of proton pump inhibitor in treating Barrett's esophagus? Am J Gastroenterol 1995;90:845-846.

14. Dubois A. Gastric emptying of liquids should not be studies independently from gastric secretion. In: Chen WY, ed. Functional disorders of the digestive tract. New York: Raven Press, 1983:151-155.

15. Scarpignato C. Pharmacologie integrée de l'estomac. In: Lewin MJM, Mignon M, eds. La muqueuse gastrique et ses sécrétions. Malesherbes: Maury-Imprimeur SA, 1985:155-175.

16. Mulholland MW, Reid BJ, Levine DS, Rubin CE. Elevated gastric acid secretion in patients with Barrett's metaplastic epithelium. Dig Dis Sci 1989;34:1329-1334.

17. Collen MJ, Johnson DA, Sheridan MJ. Basal acid output and gastric acid hypersecretion in gastroesophageal reflux disease. Correlation with ranitidine thera. Dig Dis Sci 1994;39:410-417.

18. Stein HJ, Hoeft S, DeMeester TR. Functional foregut abnormalities in Barrett's esophagus. J Thorac Cardiovasc Surg 1993;105:107-111.

19. Hirschowitz BI. A critical analysis with appropriate controls of gastric acid and pepsin secretion in clinical esophagitis. Gastroenterology 1991;101:1149-1158.

20. Hirschowitz BI. Gastric acid and pepsin secretion in patients with Barrett's esophagus and appropriate controls. Dig Dis Sci 1996;7:1384-1391.

21. DeMeester TR, Attwood SE, Smyrk TC, Therkildsen DH, Hinder RA. Surgical therapy in Barrett's esophagus. Ann Surg 1990;212:528-540.

22. Berstad A. Effect of storage on peptic activity of human gastric juice. Scand J Gastroenterol 1970;5:525-527.

23. Gotley DC, Morgan AP, Ball D, Owen RW, Cooper MJ. Composition of gastro-oesophageal refluxate. Gut 1991;32:1093-1099.

24. Savarino V, Mela GS, Zentilin P, Mele MR, Mansi C, Remagnino AC, Vigneri S, Malesci A, Belicchi M, Lapertosa G, Celle G. Time pattern of gastric acidity in Barrett's esophagus. Dig Dis Sci 1996;41:1379-1383.

25. Fein M, Fuchs KH, Freys SM. Is gastric secretion different in Barrett patients and in patients with reflux esophagitis? In: Giuli R, Siewert JR, Couturier D, Scarpignato C, eds. Barrett's esophagus. 250 questions, 250 answers. Paris: John Libbey Eurotext, 2003:179-182.

26. Calam J. Helicobacter pylori modulation of gastric acid. Yale J Biol Med 1999;72:195-202.

27. McColl KE, el-Omar E, Gillen D. Helicobacter pylori gastritis and gastric physiology. Gastroenterol Clin North Am 2000;29:687-703.

28. Smout AJ. Is the sensitivity to gastric acid inhibition Helicobacter pylori status-dependent? Scand J Gastroenterol 1998;33:32-35.

29. Scarpignato C, Pelosini I. Antisecretory drugs for eradication of Helicobacter pylori: antibacterial activity and synergism with antimicrobial agents. Progr Basic Clin Pharmacol 1999;11:135-178.

30. Dent J. Review article: is Helicobacter pylori relevant in the management of reflux disease? Aliment Pharmacol Ther 2001;15:16-21.

31. Sharma P, Vakil N. Review article: Helicobacter pylori and reflux disease. Aliment Pharmacol Ther 2003;17:297-305.

32. Scarpignato C. Gastric emptying in gastroesophageal reflux disease and other functional esophageal disorders. Front Gastrointest Res 1994;22:223-259.

33. Gillson EW, Capper WM, Airth GR, Gibson MJ, Bradford I. Hiatus hernia and heartburn. Gut 1969;10:609-613.

34. Lilly JR, Randolph JG. Hiatal hernia and gastroesophageal reflux in infants and children. J Thorac Cardiovasc Surg 1968;55:42-54.

35. Heading RC, Bolondi L, Camilleri M, Corinaldesi R, Horowitz M, Jian R, Scarpignato C. Working team report: gastric emptying. Gastroenterology Int 1992;5:203-215.

36. DeMeester TR, Johnson LF, Joseph GJ, Toscano MS, Hall AW, Skinner DB. Patterns of gastroesophageal reflux in health and disease. Ann Surg 1976;184:459-470

37. Jian R. Reflux gastro-oesophagien et motricité gastrique. Gastroentérol Clin Biol 1986;10:319-321.

38. Dubois A. Pathophysiology of gastroesophageal reflux disease: role of gastric factors. In: Castell DO, ed. The esophagus. Boston: Little, Brown and Company, 1992:479-491.

39. Hinder RA, Stein HJ, Bremmer CG, DeMeester TR. Relationship of a satisfactory outcome to normalization of delayed gastric emptying after Nissen fundoplication. Ann Surg 1989;210:458-465.

40. Hogan WJ, Dodds WJ. Gastroesophageal reflux disease (reflux esophagitis). In: Sleisenger MH, Fordtran JD, eds. Gastrointestinal disease. Pathophysiology, diagnosis and treatment, 4th ed. Philadelphia: WB Saunders Co, 1989:594619.

41. Jian R, Ducrotté F, Ruskoné A, Chaussade S, Rambaud JC, Modigliani R, Rain JD, Bernier JJ. Symptomatic, radionuclide and therapeutic assessment of chronic idiopatic dyspepsia. A double-blind placebo-controlled evaluation of cisapride. Dig Dis Sci 1989;34:657-664.

42. Scarpignato C. Gastric emptying measurement in man. Front Gastrointest Res 1990;17:198-246.

43. Heading RC, Baldi F, Holloway RH, Janssens J, Jian R, McCallum RW, Richter JE, Scarpignato C, Sontag SJ, Wienbeck M. Prokinetics in the treatment of gastro-oesophageal reflux disease. Eur J Gastroenterol Hepatol 1998;10:87-93.

44. Kogan FJ, Kotler JA, Sampliner RE. Normal gastric emptying in Barrett's esophagus. Gastroenterology 1985;88:1451A.

45. Johnson DA, Winters C, Drane WE, Cattau EL Jr, Karvelis KC, Silverman ED, Spurling TJ, Chobanian SJ, Dubois A. Solid-phase gastric emptying in patients with Barrett's esophagus. Dig Dis Sci 1986;31:1217-1220.

46. Zerbib F, Bruley des Varannes S, Ropert A, Lamouliatte H, Quinton A, Galmiche JP. Proximal gastric tone in gastrooesophageal reflux disease. Eur J Gastroenterol Hepatol 1999;11:511-515.

47. Penagini R, Hebbard G, Horowitz M, Dent J, Bermingham H, Jones K, Holloway RH. Motor function of the proximal stomach and visceral perception in gastro-oesophageal reflux disease. Gut 1998;42:251-257.

48. Stacher G, Lenglinger J, Bergmann H, Schneider C, Hoffmann M, Wolf G, Stacher-Janotta G. Gastric emptying: a contributory factor in gastro-oesophageal reflux activity? Gut 2000;47:661-666.

49. Buttar NS, Falk GW. Pathogenesis of gastroesophageal reflux and Barrett esophagus. Mayo Clin Proc 2001;76:226-234.

50. Barlow AP, DeMeester TR, Ball CS, Eypasch EP. The significance of the gastric secretory state in gastroesophageal reflux disease. Arch Surg 1989;124:937-940.

51. Boesby S. Relationship between gastro-oesophageal acid reflux, basal gastro-oesophageal sphincter pressure, and gastric acid secretion. Scand J Gastroenterol 1977;12:547-551.

52. Strader DB, Benjamin SB, Orbuch M, Lubensky TA, Gibril F, Weber C, Fishbeyn VA, Jensen RT, Metz DC. Esophageal function and occurrence of Barrett's esophagus in Zollinger-Ellison syndrome. Digestion 1995;56:347-356.

53. Scarpignato C. Acid-lowering drugs in the treatment of gastroesophageal reflux disease: an overview. Front Gastrointest Res 1992;20:129-152.

54. Ramakrishnan A, Katz PO. Pharmacologic management of gastroesophageal reflux disease. Curr Gastroenterol Rep 2002;4:218-24.

55. de Caestecker J. Prokinetics and reflux: a promise unfulfilled. Eur J Gastroenterol Hepatol 2002;14:5-7.


Publication date: August 2003 OESO©2015