1

Transactions Roy Soc Trop Med Hyg 2008; 102: 1171-1175.

Controversies in the Helicobacter pylori/duodenal ulcer story.

Michael HobsleyFrank I Tovey

Emeritus Professor of SurgeryEmeritus Senior Research Associate

Department of Surgery

and

John Holton

Reader,

Centre for Infectious Diseases and International Health

Royal Free and UniversityCollegeLondonMedicalSchool

UniversityCollegeLondon

Corresponding Author: Michael Hobsley

Fieldside, Barnet Lane

Totteridge

London N20 8AS

(+44)(0)208 4456507

Keywords:duodenal ulcer, aetiology

duodenal ulcer and diet

helicobacter pylori

gastric acid secretion

helicobacter pylori antibodies

Running title:h.pylori and duodenal ulcer: controversies

Controversies in the Helicobacter pylori/duodenal ulcer story.

Summary

In patients with Helicobacter pylori- (H pylori)-positive duodenal ulcer (DU) the organism must be eradicated to achieve rapid, stable healing. However, evidence is against much else that is commonly accepted.

1. Does H pylori cause the ulcer? Evidence against includes archaeo-pathology, geographical prevalence, temporal relationships, and H pylori-negative DU patients. DU can recur after eradication of H pylori infection; DUs may remain healed after reduction of acid secretion despite persistent infection. The faster healing of ulcers when H pylori has been eradicated is due to the organism’s interference with neoangiogenesis and the healing of wounded epithelial cells.

2.Does H pylori infection persist until pharmacologically eradicated? Studies based on current infection show that H pylori infection is a labile state that can change in three months. High rates of gastric acid secretion result in spontaneous cure, low rates permit re-infection. Hydrochloric acid, necessary for producing a DU, is strongly associated with the chance of an ulcer. At the start, patients owe their ulcer to gastric hypersecretion of hydrochloric acid; about 60% may be H pylori-negative. If acid is suppressed, the less acid milieu encourages invasion by H pylori, especially if the strain is virulent.

Text

Aetiology of Duodenal Ulcer: the Acid Era

For most of the 20th century, the orthodox opinion about the principal cause of peptic ulcer was that the stomach, when stimulated, secreted hydrochloric acid in excessive quantity. In particular, this effect was considered to be greater in duodenal ulcer (DU) rather than gastric ulcer.

This position was built up through many years of experimentation. At first, the stimulus to secretion was a ‘test’ meal of gruel, but the difficulties of separating any gruel left in the stomach from the gastric juice secreted into the stomach resulted in a search for a suitable parenteral stimulus. Finally, intravenous histamine, as the acid phosphate, delivered by a continuous infusion, seemed to be the most suitable. There accrued evidence that it could stimulate a maximal acid response, but its side-effects (flushing, headache, etc.) were unpleasant, and an antihistamine (H1 antagonist) had to be used to allow a sufficient dose to be given so as to achieve the maximal response. Then a variety of errors affecting the collection of the gastric juice from the stomach had to be addressed. These errors included losses from the stomach into the duodenum (minimized, but not prevented, by continuous gastric aspiration), reflux from the duodenum of alkaline contents into the stomach, and spontaneous compartment formation in the stomach into pouches, probably provoked by continuous suction.

Methods were developed for correcting these errors: they relied on the slow infiltration down a second channel of an inert marker such as the dye phenol red, the measurement of electrolytes, and corrections for pyloric loss and duodenogastric reflux (Whitfield and Hobsley, 1979). As a result of the use of such methods and the proof of their accuracy, it became established that the maximal secretion of hydrochloric acid by the stomach was dependent on stature, and that the best (and simplest) correlate with stature was the subject’s height.

Earlier work had established that patients with duodenal ulcer (DU) lay in two groups related to their maximal secretory capacity: those lying above the 95% tolerance limits of the normal population, and those lying within the normal limits. Thus, the distinction between large and smaller amounts of acid meant that some other cause must prevail in the normal secretory group. However, no exception was found to the original dictum of Schwarz (Schwarz 1910), No Acid, No Ulcer. Indeed, there was at the lower border of the range in non-ulcer individuals a band of secretion in which no subject with peptic ulceration lay.

The final result of making all the corrections to the measurement of gastric acid secretion was reported in 1987 (Hobsley and Whitfield, 1987). They made no difference to the broad picture described in the previous paragraph: patients with DU never had an abnormally low secretion when maximally stimulated, but only some of them had a supra-maximal secretion. However, there was one important further element: in each band of arbitrarily designated secretion in the normal range, the risk of developing a duodenal ulcer increased with increasing maximal acid secretion, until, at greater than the 95% upper tolerance limit of the population, it became 100%. Moreover, Helicobacter pylori (H pylori)-status was not assessed in these studies. This last point is important and will be referred to again later.

Aetiology of Duodenal Ulcer: Rôle of Helicobacter pylori

The award of the Nobel Prize in medicine or physiology to Warren and Marshall for the discovery of H pylori recognized the importance of eradicating the organism in patients with duodenal ulcer, a procedure that converts a chronic relapsing situation into an acute and rapidly curable one. The impression seems to have been widely accepted that H. pylori actually causes the ulcer and that there is little more of interest to be discovered about the condition: for example, the journal Gut no longer has a section of its original research communications devoted to the subject of peptic ulcer. However, we feel that this view is too simplistic, and that much remains to be learned in the field of acid and duodenal ulcer. Papers describing measurements of gastric secretion have virtually disappeared from the medical literature. In particular, there are three important controversies in current thinking:-

1)That Helicobacter pylori infection causes duodenal ulceration

2)That virulent strains of the organism are more likely to cause duodenal ulceration than non-virulent strains.

3) That H pylori infection is a chronic infection that remains with the patient indefinitely (in at least most cases) unless eradicated by pharmacological means.

Controversy 1.The presence of H pylori causes duodenal ulcer.

The evidence for this causal link is that the prevalence of H pylori infection in DU patients is higher than in controls, a condition noted in all studies, whether in areas of high or of low population prevalence of H pylori, and that the treatment of H pylori infection leads to healing.

Against this concept that H pylori infection causes ulceration are the following points:-

a)There is evidence that H pylori infection has been present in humankind for centuries (Falush et al. 2003, Linz et al. 2007,Yamaoka et al. 2000, 2002; but although heartburn and indigestion have been described for thousands of years, gastric ulcers were identified only in the 18th century and duodenal ulcers only in the nineteenth (Baron, 2000). Indeed, duodenal ulcers only became common at the beginning of the 20thCentury (Kidd and Modlin, 1998; Baron and Sonnenberg 2001, 2002). No convincing explanation can be offered for various features such as the fact that peptic ulcer disease seems to have begun around 1750, mainly in the gastric rather than the duodenal region and affecting predominantly young women; the emphasis did not change to mainly duodenal ulcer predominantly in men until the beginning of the twentieth century.

b) The prevalence of duodenal ulceration is not higher in countries with a high

prevalence of H pylori infection as would be expected if it were causal (Hobsley

and Tovey, 2001; Hobsley at al. 2006).

c)Geographically the prevalence of duodenal ulceration does not correspond with the prevalence of H pylori infection (Hobsley and Tovey, 2001; Hobsley et al. 2006). Within countries with the same overall prevalence of H pylori infection the prevalence of duodenal ulceration may vary from region to region. This is borne out particularly in studies from India (Tovey et al. 2004), Africa (Holcombe,1992; Segal et al. 1998, 2001; Tovey et al. 2005) and China (Wong et al.1998; Ching and Lam, 1994).

d) There is considerable evidence of duodenal ulceration occurring in the absence of H

pylori infection:-

i)There are only three reports (Boulos et al. 2002; Pest et al. 1996; Bytzer and Teglbjaerg, 2001) in the literature of H pylori status of cases presenting with a short history of duodenal ulceration (under 6 months) compared with those with a longer history. Without exception, these reports agree that patients with the short history are less likely to be H pylori-positive than those with a longer history. If one accepts that a cause must precede an effect, this is a remarkably contrary finding.

ii) There are, from different countries, a large number of endoscopy reports

of H pylori-negative duodenal ulceration unrelated to NSAIDs or other

morbid conditions, ranging from 14% to 72%, and occurring more

often in countries with a low prevalence of H pylori infection.(Hobsley and

Tovey 2001; Hobsley et al. 2006;Laine et al. 1998; Jyotheeswaram et

al.1998; Borody et al. 1991; Ciociola et al. 1999. If the organism is

responsible for causing duodenal ulcer, we need to find a cause for the patients who have an ulcer but are not infected

.

e) Duodenal ulceration can recur after eradication of H pylori infection (without re-

infection).(Hobsley and Tovey, 2001; Miwa et al.2004; Martino et al. 1999; Forbes

et al.1994)

f) Duodenal ulcers may remain healed after medical or surgical reduction of acid

secretion despite persistence of infection with H pylori.(Hobsley et al, 2006;

Bytzer et al, 2000; Prach et al. 1998; Martin et al, 1995; Huang et al.2004)

g)Patients with duodenal ulcer who are infected have a smaller maximal gastric

secretion than those who are not infected (Chandrakumaran et al. 1994). In the light of the secretion studies that confirmed increased maximal gastric secretion as a major concomitant of duodenal ulcer, this is a disturbing finding for those who believe that the organism causes duodenal ulcer by causing hypersecretion in response to hypergastrinaemia. but only to be expected as the organism is known to cause gastritis.

Controversy 2:- Virulent strains of H pylori are more likely to cause duodenal ulceration.

This contention is based on the finding that duodenal ulcer patients are more likely to be infected with virulent strains than the normal population. Against the concept that the virulent strains are more likely to be the cause of duodenal ulceration is the fact that the prevalence of duodenal ulceration in a country is not related to the prevalence of virulent strains. In countries where the prevalence of virulent strains is high there is no corresponding increase in the prevalence of duodenal ulceration (Tovey et al. 2006) Moreover, there is a simple alternative explanation for the finding (see later).

Controversy 3:-H pylori infection is chronic and static.

It is commonly thought that H pylori infection of the gastroduodenum is static: once present, the organism has to be eliminated by pharmacological treatment. This idea has become accepted on the basis of serological evidence that is hardly convincing: seropositivity is not synonymous with current infection, nor is seronegativity with absence of current infection. On the contrary, there is convincing evidence that infection is a labile state: Individuals may move between infected and non-infected states at intervals as frequent as 3 months.(Hobsley et al2007) The deciding factor appears to be the rate at which the subject’s gastric hydrochloric acid is being secreted (see below) : high secretion rates result in spontaneous cure, but if the secretion rate falls somewhat re-infection can occur.

If H pylori infection is not static or causal of duodenal ulceration, what is its role?

We suggest that in the presence of an ulcer H pylori infection acts in a similar way to a secondary infection in a wound, causing delay in healing and chronicity, and that virulent strains are more likely to have this effect. There is evidence that H pylori infection interferes with neoangiogenesis (Pearce et al.2004) and in the process of healing of wounded duodenal epithelial cells (Tabel et al. 2003). This would explain the high prevalence of H pylori infection in patients with persistent duodenal ulcers, and why treatment and eradication results in rapid and more stable healing.

The role of acid in duodenal ulceration and in H pylori infection.

As indicated in the introduction, the chance of an individual developing a duodenal ulcer is strongly statistically related to the rate at which the stomach secretes acid when maximally stimulated. This result was obtained after correction for height, collection errors such as pyloric losses and duodenogastric reflux, and without determining H pylori status.(Hobsley and Whitfield, 1987)

H pylori infection is not chronic but labile, depending on local acid concentration. When Marshall infected himself with the organism he suppressed his acid secretion before swallowing the culture. Two weeks later, when acid secretion was normal, the infection had disappeared spontaneously (Marshall et al. 1985).

In vitro, growth of H pylorI is restricted to pH levels of 6.5 - 7.5, the optimal pH being 7.0. At pH 3.0 - 3.5 the organism is no longer viable unless urea is added. H pylori produces urease and this reacts with urea producing ammonia. With the addition of urea, growth can occur down to a pH of 1.5, below which the organism is not viable.

Colonization of the stomach by H pylori depends on favourable pH levels: both highly acid and highly alkaline conditions kill the organism, the range permitting growth being between pH 1.5 and pH 7.5.(Sjostrom and Larsson, 1996; Marshall et al. 1990; Clyne et al.1995) Thus the absence of H pylori infection in early cases of duodenal ulceration may be explained by the initial high levels of acid secretion. The presence of Hpylori in later cases may be the result of treatment and the reduction of acid secretion, permitting colonization with Hpylori. It is in this process of re-colonization that virulent strains may outperform non-virulent ones, so accounting for the observed preponderance of virulent strains in patients with duodenal ulcer.

The higher prevalence of H pylori-negative DU in countries with a low prevalence of H pylori infection is mathematically explicable with a model in which most patients

(about 60%) with an early DU are H pylori-negative, but become positive as time passes (about six months) (Hobsley et al. 2007). This last point in particular fits exactly with the concept that high acid secretion causes the DU and prevents infection with H pylori, but that later treatment with acid-diminishing drugs reduces gastric acid secretion and thus encourages infection. This hypothesis correctly predicts the relationship between H pylori –positive and H pylori-negative patients with duodenal ulcer, both in countries where the prevalence of the organism is high (about 80%), and where it is low (about 40 %)

Apart from NSAIDs and co-morbid conditions (e.g. Crohn`s disease), our conclusion is that acid is the main factor in the cause of duodenal ulceration. Possibly it is the only necessary factor in hypersecretors; but in moderate secretors there may be other factors affecting mucosal defence. Smoking increases gastric secretion as a chronic effect, although during smoking gastric acid secretion is suppressed (Whitfield and Hobsley, 1987; Roxburgh et al. 1992). In addition there are dietary protective factors in certain staple foods, a fact for which there is strong geographical evidence (Tovey et al. 1989, 2004, 2005; Wong et al. 1998; Tovey, 1992, 1994; Jayaraj et al. 1980; Tovey and Tunstall, 1975). There is experimental evidence from animal models that these foods enhance mucosal protection against ulceration. This protective effect is related to particular lipids present in the protective foods. The lipid extract from these foods, and particularly the phospholipid and sterol fractions present in the extract (Jayaraj et al. 2000, 2003) give protection against ulceration, Where these lipids are present in the staple diet the geographical prevalence of duodenal ulceration is low and where they are absent the prevalence is high, irrespective of the prevalence of H pylori infection.

Finally one must bear in mind the evidence from the acid secretion studies that whatever is the factor causing ulceration, it appears to be, in the statistical sense, random. Within the range of secretion of the normal population, the chance of developing a duodenal ulcer increases smoothly with rate of maximal secretion of acid, but in the end mucosal resistance to acid may be the deciding factor in the development of an ulcer.

Besides the factors of smoking or the presence of dietary protective lipids affecting mucosal resistance, is it possible that we are also dealing here with a traumatic factor? The most likely vehicle for such a factor (considering that the common areas for peptic ulcer are the lesser curvature in the stomach and the point in the duodenum where the stream of chyme injected into the duodenum by the pyloric muscle hits the duodenal wall) is ingested food.

Our conclusion is that in the presence of duodenal ulceration due to the above factors, Hpylori infection, whether primary or secondary, results in a chronic relapsing situation and treatment of the infection leads to a rapid and more stable cure. If we want to learn more about such factors, we will do well to avoid assuming that we know all we need to know about the aetiology of duodenal ulcer. Moreover, this subject has an impact on the idea that it may be good to eliminate the organism from whole populations. If our analysis (Hobsley, Tovey and Holton, 2006) is correct, 60% of duodenal ulcer patients at the inception of the disease are free from the infection and therefore make more acid than do H pylori-negative subjects, such a policy would result, in developed countries, in a modest increase in prevalence of duodenal ulcer of 30%, but in developing countries the corresponding increase would be threefold.

Funding:None

Conflicts of interest:None declared

Ethical approval:Not required

References;

Baron J.H., 2000. Peptic ulcer. Mt Sinai J. Med., 67: 58-62.

Baron J.H., Sonnenberg A. 2001. Hospital admissions for peptic ulcer and indigestion in London and New York in the 19th and early 20th centuries. Gut, 50: 568-570

Baron J.H., Sonnenberg A. 2002. Alimentary diseases in the poor and middle class in London 1773-1815, and in New York poor 1797-1818. Aliment. Pharmacol. Ther. 16(10): 1709-1714.