LETTER

Besides the gastric bacterium Helicobacter pylori, strains of Pseudomonas, Campylobacter, Bacteroides, Bifidobacterium Eubacterium, and Staphylococcus-like bacteria can be found in the human stomach in a state of disease (1, 2). Pseudomonas aeruginosa infection can lead to false-positive results of immunological tests to detect H. pylori (3) and may inhibit its in vitro growth (4). Contaminants or gastric non-H. pylori bacteria can grow in cultures of gastric specimens. We use brucella agar plates containing 5% calf serum, vancomycin (10 mg/ml), polymyxin B (10 mg/ml), and nalidixic acid (20 mg/ml) for culturing H. pylori from gastric biopsy specimens. We have found pure cultures of gram-negative bacilli that were positive for rapid urease, catalase, and oxidase tests but with colony characteristics (white small colonies rather than the transparent ones typical of H. pylori) that made us suspect that the strains were not H. pylori. We decided to further investigate three of these strains (isolated from three different patients) that could have lead to false identification of H. pylori, if only cell morphology and biochemical tests would have been considered. Unlike H. pylori, our three isolates could grow aerobically. Electron microscopy observations showed gram-negative bacilli with multiple unipolar flagella, also found in H. pylori. Two isolates had a unipolar tuft of two or more flagella, and the other had one polar flagellum. Analysis of the cell membrane fatty acids of the three strains (Midi System; Hewlett Packard) indicated a high index of identity with Pseudomonas aeruginosa and the organisms formerly known as Pseudomonas cepacia and Pseudomonas diminuta (currently known as Burkholderia cepacia and Brevundimonas diminuta, respectively). All three species have been previously encountered in clinical specimens. Pseudomonas species may have a polar monotrichous flagellum (P. diminuta) or fewer than three flagella (P. aeruginosa [96%]) or have a tuft of three to eight flagella (P. cepacia [99%]). They are catalase-positive and can also be positive for urease (P. aeruginosa [72%]), P. cepacia [46%]), and P. diminuta [<10%]), oxidase, and resistant to polymyxin (P. aeruginosa [3%]), P. cepacia [>90%]), and P. diminuta [10%]).

To confirm H. pylori identify, colony morphology is very important, and testing strict microaerophylism would resolve dubious identifications. Since culturing of H. pylori is being performed in an increasing number of clinical and research laboratories, it must be stressed that relying exclusively on cell morphology and positive reactions for rapid urease, catalase, and oxidase tests for confirmation of H. pylori may be misleading.