Background-The development of macrolide resistance in Helicobacter pylori is considered an essential reason for failure of antibiotic eradication therapies. The predominant mechanism of resistance to macrolides, particularly clarithromycin, is based on three defined mutations within 23S rRNA, resulting in decreased binding of the antibiotic to the bacterial ribosome. Aim-To develop an rRNA based whole cell hybridisation method to detect Helicobacter species in situ within gastric tissue, simultaneously with its clarithromycin resistance genotype. Methods-A set of fluorescent labelled oligonucleotide probes was developed, binding either to H pylori 16S rRNA or 23S rRNA sequences containing specific point mutations responsible for clarithromycin resistance. After hybridisation and stringent washing procedures, labelling of intact single bacteria was monitored by fluorescence microscopy. The new approach was compared with PCR based assays, histology, and microbiological culture. Results-In comparison with the phenotypic resistance measurement by E test, the genotypic clarithromycin resistance correlated perfectly (100%) for 35 H pylori isolates analysed. In a set of gastric biopsy specimens (27) H pylori infection was confirmed by histology (17/27) and correctly detected by whole cell hybridisation. Five clarithromycin resistant strains were identified in gastric tissue specimens directly. Furthermore, non-cultivable coccoid forms of H pylori were easily detectable by whole cell hybridisation. Conclusions-Whole cell hybridisation of rRNA holds great promise for cultivation independent, reliable, and rapid (three hours) genotypic determination of clarithromycin resistance in H pylori. Compared with PCR techniques it is independent of nucleic acid preparations, not prone to inhibition, and allows semiquantitative visualisation of the bacteria within intact tissue samples. (Gut 2000;46:608-614)

In a study to determine secondary resistance among Helicobacter pylori isolates, gastroenterologists from several German cities submitted over a 3-year period to centre A (Regensburg) or centre B (Freiburg) gastric biopsies from patients in whom one or more therapies to eradicate Helicobacter pylori had failed. Rates of resistance among the collections of 302 (centre A) and 252 (centre B) isolates were, respectively, as follows: to metronidazole, 75% and 66%; to clarithromycin, 58% and 49%; to amoxicillin, 0%; to ciprofloxacin, 9%; to doxycycline, 0%; and to rifampin, 0%. Resistance to clarithromycin was associated with metronidazole resistance in 89% and 85% of the isolates in centre A and centre B, respectively.

The vacuolating cytotoxin of Helicobacter pylori (VacA) is known to cause cell damage to mammalian cells and is suspected to give rise to gastric epithelial lesions that might lead to peptic ulcer disease. As shown recently, the gene encoding VacA exhibits genetic variation, with three different families of signal sequences (s1a, s1b, and s2) and two families of midregion sequences (m1 and m2). In order to investigate the relationship between the presence of specificvacA genotypes and peptic ulceration, the vacAgenotypes of 158 clinical isolates of H. pylori were determined. The study group consisted of 106 patients with duodenal ulceration; 52 patients with nonulcer dyspepsia (NUD) were used as controls. H. pylori of genotype s1 was isolated from 96% of the patients with ulcerations, whereas genotype s2 was only present in 4%, indicating a strong correlation between thevacA genotype and peptic ulceration (P < 0.001). In contrast, 31% of the patients from the NUD control group were infected with strains of vacA genotype s2. Particular midregion genotypes (m1 and m2) were not associated with clinical manifestations. The midregions from 18% of the isolates could not be classified by the proposed scheme. DNA sequencing revealed high homology between the untypeable midregions and that of genotype m1, with multiple base pair exchanges, some affecting the primer annealing site. Compared to those of m1 and m2 alleles, the divergent midregions from untypeable strains showed clustering, indicating the presence of a further subfamily of sequences in the midregion of vacA in German isolates, for which we propose the term “m1a.” A new specific primer that we designed for typing m1a isolates might be useful in other studies.

The two pantoprazole-based triple therapies tested in this study are equally effective in H. pylori eradication, ulcer healing and relief from ulcer pain. It is concluded that 7 days of triple therapy are generally sufficient.