Because it has been common cause since its discovery in duodenal ulcer (DU) patients by Brian Marshall in 1982. Initially it was shown that 90% of duodenal ulcers are associated with Helicobacter pylori (HP) infection. Furthermore subtypes were identified that were associated with DUs and other subtypes with gastric cancer. The jury was in and the onslaught justified. The development of simple blood tests, in addition to stool and saliva tests simplified the decision-making so that all patients with dyspepsia were tested and if found positive treated with clarithromycin and amoxicillin. A regimen that will clear 80 to 90% of infections but resistance is growing.
This acceptance of reflex eradication was based on the cost and invasiveness of a gastroscopy needed to make an accurate diagnosis as compared to the relative cheapness of the antibiotics and the hope that the risk gastric cancer might be reduced. The use of antibiotics also avoided the need for expensive proton pump inhibitors. Recently however, these have fallen faster than the cost of the antibiotics so that a month’s course of PPI is now less than 7 days clarithromycin.
The first fallacy in this protocol was the confusion between sensitivity and specificity. Whereas originally 90% of duodenal ulcers were associated with HP infection, this rate has been falling progressively in the Western world and is now closer to 50%. More important, however, is the specificity. At best only 15% of patients with HP have ulcer disease and only 40% of patients with dyspepsia have a duodenal ulcer. The latter has also been falling as the incidence of oesophageal reflux has been increasing. Furthermore, other conditions which require confirmatory biopsy such as gastric ulcers and gastric cancer will be missed or the diagnosis delayed during the pointless eradication of HP. In fact, almost half the patients with a positive HP test have no gastro-duodenal pathology at all. Some HP strains appear to protect against peptic ulcer disease but methods to differentiate the good from the bad are not widely available. In this context the side effects of clarithromycin need to be remembered. Gastrointestinal symptoms occur in up to 7% of patients. More important is the very rare but potentially fatal liver failure. Furthermore drug interactions are assured by its inhibition of the CYP3A4 pathway.
The incidence of oesophageal disease such as Barrett’s metaplasia and upper oesophageal cancer is associated with the absence of gastric HP. In fact it could be interpreted that the presence of HP protects against the development of these GORD complications. More surprising is the recent observation that childhood asthma, allergic rhinitis, and atopy are more common in HP negative children. This is even more intriguing as the prevalence of HP in children younger than 10 years in the Western world has decreased from 70%-90% previously to less than 10%. As in Crohn’s disease it might be that a change in the natural flaura and fauna of the gastrointestinal tract has unintended consequences on gastro-duodenal and intestinal mucosal defences.
So where do we stand in the HP / DU debate.
On the negative side
On a more philosophic note it appears that HP arose in ancient Africa 60 000 years ago and has changed/developed as it spread around the world with early man. The incidence is now falling probably due to better hygiene and widespread use of antibiotics. The incidence of DU, particularly in men, rose in the 20 century but has been falling for the past 25 years. It seems to have been replaced with oesophageal disease which has an inverse relationship with HP infection.
The days of reflex eradication of HP seem to be numbered…