Community acquired acute Helicobacter pylori gastritis

See article on page 1377 Nomura et al. reported results of a prospective study of community acquired acute Helicobacter pylori gastritis. 1 Patients underwent endoscopy because of a recent onset of upper gastrointestinal symptoms. The diagnosis of acute H. pylori gastritis was suspected based upon endoscopic findings of acute mucosal damage (e.g. multiple erosions, ulcers and friability). Histologic findings showed acute inflammation without a significant chronic inflammatory reaction. In each case H. pylori infection was confirmed by the use of a positive rapid urease test, as well as by culture and histology. Acute H. pylori infection was also suspected because of the presence of infection in the absence of a postive immunoglobulin (Ig)G serology for H. pylori antibodies. The authors prospectively divided the patients into two groups; those who presented within 24 h of symptom onset and those who presented 4–6 days after symptoms appeared. Those with an acute presentation were randomized to receive either antimicrobial therapy with lansoprazole or lansoprazole alone. All patients who presented after prolonged symptoms received antimicrobial and antisecretory therapy. There was a rapid resolution of symptoms following the institution of therapy. The authors attributed the response to the therapy but as all patients received lansoprazole it is impossible to exclude spontaneous remission. Treatment with lansoprazole plus antibiotics resulted in a resolution of the endoscopic finding within 1 week. However, lansoprazole alone did not accelerate the healing of the endoscopic damage, because those receiving lansoprazole alone had a similar course to those who presented with symptoms several days before receiving antibiotic therapy. All patients had normal endoscopy by the third week. 1 Four of the five patients who received only lansoprazole had a ‘spontaneous’ clearing of the infection. Because one of the five patients (20%, 95% CI = 1–72%) did not have a clearing of the infection, the authors suggested that the diagnosis of acute H. pylori infection should prompt antibiotic therapy to ensure H. pylori eradication and prevention of chronic H. pylori gastritis. Proton pump inhibitors have potent anti-H. pylori activity and their use is associated with a marked fall in bacterial load; more than 20% become negative by histology or urea breath test (UBT). The anti-H. pylori activity of the proton pump inhibitor casts some doubt on the interpretation of the 80% (e.g. four of five) ‘spontaneous remission’ rate reported by Nomura et al.1-4 When acute H. pylori gastritis involves the gastric corpus, marked hypochlorhydria or achlorhydria is expected. 5-10 Proton pump inhibitors would not be expected to be beneficial, based on antisecretory activity. It is therefore unclear whether the symptomatic response reported was related to the therapy given, the natural history of the disease, or both. The presumption of the authors was that those patients with one day symptoms would continue to have symptoms and be like the group that had them for 4–6 days. The different pattern and duration of symptoms in the prospectively infected individuals suggests that that presumption might not have been valid. Although there are a number of publications, very little is known about the acute phase of H. pylori gastritis. 5-26 This study shows that the prepared mind can prospectively identify some cases. Nevertheless, the spectrum of the initial presentation remains unknown. The presentation may depend on the inoculum size, the virulence of the infecting strain of H. pylori, and/or undetermined host factors. The endoscopic findings in this study included multiple erosions and or ulcers with friable mucosa. These findings are suggestive of, but not specific for acute H. pylori gastritis. The authors actually evaluated 18 patients with a similar presentation and excluded eight patients (44%) because of evidence of chronic H. pylori infection (anti-H. pylori IgG positive in seven) or negative culture and histology for H. pylori (in one patient). In our experience with feeding H. pylori to normal volunteers, approximately 70% developed dyspepsia, whereas no patient had abnormal gastric mucosa after 2 weeks, despite them not receiving any therapy. 27 We used a cagA negative strain and yet all our patients had an intense polymorphonuclear infiltrate in the gastric mucosa. Whether cagA status is a critical determinant with respect to the endoscopic finding is unknown. Eventually, prospective challenge studies using cagA positive and negative strains will provide a better insight with regards to the spectrum of the initial manifestations than are possible from studies in the community. In the USA, it would be unlikely that any of the patients presenting with the symptoms of acute H. pylori gastritis would have received an endoscopic evaluation. The criteria for the presumptive diagnosis of acute H. pylori infection do not require endoscopy, but do require objective evidence of H. pylori infection (e.g. [13C]-UBT positivity) and an absence of serum IgG antibodies to H. pylori. Most of the patients in the USA and Japan who presented symptoms would most likely have received symptomatic therapy with an antisecretory drug. The data from the Nomura et al. study suggest that a spontaneous cure may be common. If a spontaneous cure or cure related to the use of a proton pump inhibitor occurs commonly, it is unlikely that the correct diagnosis would be considered or confirmed. In our local experience with children, over a 5-month period, 52 children (aged less than 13 years) who presented with acute upper gastrointestinal symptoms had a [13C]-UBT and H. pylori IgG serology. Twelve (21%) had a positive [13C]-UBT, but only three had positive serology, suggesting that the results from Japan are not unique and deserve wider appreciation (MA Gilger and AR Opekun, unpubl. obs., 2000). As with most studies, new information brings up as many questions as it answers. Even when one is looking for acute H. pylori, it is uncommon even in a country with a high background prevalence of the infection. One presumes that only the tip of the iceberg was sampled. For example, why did these particular individuals decide to come to the doctor after having symptoms for only one day? How many patients who had symptoms for one day would have had symptoms at 4–6 days? Do symptoms correlate with the extent, severity or typographic distribution of endoscopic or histologic damage? Where did they acquire the infection? Were there any common characteristics in the week or two prior to presentation? Was the infecting strain the same as that of their spouse? One can list many more questions illustrating how little we know and how much there is yet to learn. In conclusion, one should suspect acute H. pylori when acute gastric mucosal damage (erosions and ulcers) is not otherwise expected (e.g. chronic non-steroidal anti-inflammatory drug (NSAID) use). A positive rapid urease test would clinch the diagnosis of H. pylori, but would not differentiate between acute H. pylori, chronic H. pylori and other processes such as syphilis, NSAID damage or an unknown disease. 28 Nevertheless, most agree that the diagnosis of H. pylori infection should prompt therapy. The choice of therapy should be based on which combination (drugs, doses, duration) provide the optimum results. The quest for the shortest duration of therapy has resulted in an unacceptable rate of failures. 29 In my opinion, the choice should be an antibiotic combination that is proven to provide the best results in the population, taking into account the local prevalence of antibiotic resistance. 29