Evaluation of two commercial enzyme immunoassays for the diagnosis ofHelicobacter pylori infection

Sera from 65 patients with upper gastrointestinal tract symptoms and provenHelicobacter pylori infection, and from 42 negative controls were tested with two commercial EIAs (GAP test, Bio-Rad; and ECP test, Biometra) and two non-commercial EIAs, one performed with whole sonicated cells and the other with acid extract ofHelicobacter pylori as antigen. The GAP assay showed a sensitivity of 83.1 % and a specificity of 47.6 %. The ECP assay showed a sensitivity of 87.7 % and a specificity of 61.9 %. For both non-commercial EIAs these figures were 87.7 % and 88.1 %, respectively. Independent of the interpretive criteria established by the manufacturers, receiver operating characteristic curves were plotted for better evaluation of the four methods. Both commercial tests showed a lower probability of yielding a correct diagnosis than the non-commercial tests (p <0.05). Although commercial EIAs are convenient for the diagnosis ofHelicobacter pylori infection, the accuracy of the two commercial tests evaluated in this study was lower compared to that of the two non-commercial EIAs.     

Rapid growth ofHelicobacter pylori

The ability of six variants of a new charcoal medium and Skirrow's medium to growHelicobacter pylori in 3 and 5 days was studied using 20 different strains ofHelicobacter pylori. The main admixtures for the charcoal media were serum, whole blood, and egg yolk emulsion. For this purpose, serum was significantly better and egg yolk emulsion significantly worse than whole blood. The addition of IsoVitalex resulted in significantly improved growth on the charcoal media. Skirrow's medium showed very poor performance after three days of incubation and needed a long incubation time.     

Evaluation of a serological test for diagnosis ofHelicobacter pylori infection in children

A serological test for the diagnosis ofHelicobacter pylori infection (Cobas Core Roche, IgG, 2nd Generation; Roche, France) was compared with the examination of biopsy samples (culture and histology) obtained after endoscopy in 115 children to assess its value. In 94 children (42 positive and 52 negative), results were concordant. In 10 children a positive serological test was associated with an absence ofHelicobacter pylori, while in 11 others a negative serological test was associated with a positive culture. Sensitivity of the test was 79.2% and specificity 83.9%. A relationship between IgG titers and age (r=0.31, p<0.05) was found. Serological tests could be useful for the diagnosis ofHelicobacter pylori infection, but a negative test result does not rule out infection, particularly in children under 10 years of age.     

Catalase negative mutants ofHelicobacter pylori

Nine strains ofHelicobacter pylori have been isolated exhibiting spontaneous mutations with a loss of catalase activity. Growth characteristics in vitro were unaffected by the mutation showing that catalase is not essential for growth ofHelicobacter pylori. Parent strains and mutants could not be distinguished morphologically from each other when compared by electron microscopy. Restriction endonuclease digestion withHindIII, separated in an 0.7 % agarose gel in TBE buffer, showed each pair to be highly related to each other. SDS-PAGE separation of proteins from four mutants and parent strains showed that all mutants lacked a 57 kDa protein. The partial N-terminal sequence of this protein shows homology with maize catalase and may be the subunit of theHelicobacter pylori catalase tetramer. It is concluded that catalase negative mutants ofHelicobacter pylori occur spontaneously in vitro, but have not yet been observed in vivo. The paucity of such catalase negative strains in clinical specimens could mean that catalase is a virulence factor in vivo that puts mutants at a selective disadvantage.     

Microbiological aspects ofHelicobacter pylori (Campylobacter pylori)

The human gastric pathogenCampylobacter pylori has recently been reclassified asHelicobacter pylori, and a related spiral bacterium found in the stomach of ferrets has been designatedHelicobacter mustelae. The general microbiological features ofHelicobacter pylori are delineated here, with details of phenotypic differences betweenHelicobacter pylori andHelicobacter mustelae; comparisons are made withWolinella succinogenes andCampylobacter jejuni. TheHelicobacter organisms possess an external glycocalyx which can be visualised by electron microscopy, and which may be involved in bacterial adherence. The finding of soluble and cell-associated haemagglutinins ofHelicobacter pylori is reported. Detection ofHelicobacter pylori in clinical specimens, susceptibility of the organism to antibacterial agents, and other aspects of practical and clinical significance are briefly reviewed.     

Cell-associated haemolytic activity ofHelicobacter pylori

Helicobacter pylori cells cultured on solid medium were quantitatively tested for haemolytic activity against erythrocytes of man, sheep, the guinea pig and rabbit. Using 4-day and 8-day cultures of two standard strains (ATCC 43504, IMMi 676), human erythrocytes were not lysed by 10 % bacterial suspensions. Rabbit erythrocytes were the most sensitive to 8-day cultures. Hot-cold incubation yielded the highest haemolysis titres. The extent of haemolysis strongly correlated with the number of bacterial cells. Supplementation of the test medium (PBS, pH 7.4) with L-cystein, dithiotreitol, MgCl₂, EDTA, cholesterol, lecithin or sphingomyelin did not influence the haemolysis titres. They were significantly reduced in the presence of pronase E, human serum, bovine serum albumin or CaCl₂, and by heat treatment of the bacteria. Supplementation of the test medium with cardiolipin strongly increased the haemolysis titres. Comparing the cell-associated haemolytic activity of 18 strains, the titres ranged from < 2 to 64, with a median titre of 16. No correlation was found between the haemolytic activity and phospholipase C activity of the cell suspensions. It was concluded that the formation of lysophosphatides and non-enzymatic factors rather than a sulphydryl-activated cytolysin or phospholipase C are responsible for the cell-associated haemolytic activity. This property may be involved in the pathogenicity and virulence ofHelicobacter pylori.     

Evaluation of a monoclonal antibody for detection ofHelicobacter pylori in a direct immunofluorescence test

A monoclonal antibody was developed for detection ofHelicobacter pylori in gastric tissue sections in a direct immunofluorescence test. On a comparison of the immunofluorescence test with standard methods for detection ofHelicobacter pylori, i.e. culture, the urease activity test and histological examination of tissue sections, using 158 biopsy specimens, 30 specimens were positive in all methods and 64 negative. In the remaining cases comparison was not possible because either immunofluorescence (29 specimens) or the standard methods (16 specimens) gave ambiguous results. The direct immunofluorescence test may have potential as an alternative to standard methods, but further testing in a defined patient population is necessary.     

Ultrastructural localization of urease ofHelicobacter pylori

Helicobacter pylori urease was characterized by means of an enzyme histochemical electron microscopic technique. Ultrastructural analysis revealed no urease activity in one strain; in sevenH. pylori strains (43.75%), urease activity was associated with the cell membrane. Eight strains (50.0%) showed reaction product located within the cytoplasm. Urease activity showed no correlation with localization of activity. Our results demonstrate thatH. pylori urease is not uniform in allH. pylori strains, and differences in activity and localization of urease activity may account for different virulence activities.     

Adaptation ofHelicobacter pylori to aerobic growth

Two of 23 strains ofHelicobacter pylori adapted from microaerobic to aerobic growth on blood agar plates incubated in humidified air. The airadapted strains remained urease and phenylalanine deaminase positive and did not require the buffering effect of an enriched CO₂ atmosphere for growth. The significance of this phenomenon remains to be determined as the two strains capable of aerobic metabolism were laboratory-adapted.     

Development of a rapid PCR method for identification ofHelicobacter pylori in dental plaque and gastric biopsy specimens

A rapid and simple polymerase chain reaction (PCR) method was developed to detectHelicobacter pylori in gastric biopsy specimens and dental plaque samples. The primers were targeted to the 16S rRNA sequence ofHelicobacter pylori strain ATCC 43504. The system was found to have a theoretical detection level of 0.5 to 5Helicobacter pylori cells in a 5 l sample of dental plaque. In the absence of plaque, the detection level was even better: theoretically, 0.05 to 0.5Helicobacter pylori cells were detected in water suspension. However, this appeared to be due to the presence of free bacterial DNA in the culture used for the sensitivity determination. Thus, the actual sensitivity of the system was found to be fewer than fiveHelicobacter pylori cells, irrespective of the type of sample used. The method was then used to analyse 29 dental plaque and gastric biopsy specimens collected from patients with a history of recurrent peptic ulcer disease. Fourteen stomach specimens were positive forHelicobacter pylori when tested with the PCR method, while the respective figures with culture, histological examination and the urease test were 11, 12 and 9. No positive dental plaque samples were observed.     

Sensitivity ofHelicobacter pylori to different bile salts

The sensitivity ofHelicobacter pylori to different unconjugated or conjugated bile acids both on BHI blood agar and in BHI broth supplemented with starch, horse serum or egg yolk was studied. Bile salts were more toxic in the medium containing starch than in the media containing horse serum or egg yolk, and unconjugated bile salts were more toxic than the conjugated salts. Deoxycholic acid was the most toxic of the bile acids studied. Ox bile was bacteriostatic at the 2 % level. Bile Salts 3 mixture was bactericidal at a concentration of 0.25 mmol.     

Evaluation of staining methods for identifying Campylobacter pylori

Campylobacter pylori has been implicated in the pathogenesis of peptide ulcer disease. The rapid identification of this organism may depend upon histologic diagnosis, because culture methods are complex and require a minimum of seven days in order to identify a negative specimen. The purpose of this study was to determine which stain used to identify this organism was the most cost-effective and easiest to perform and interpret on a routine basis. Sixty-one consecutive gastric antral biopsies were stained with hematoxylin and eosin, Giemsa, Brown-Brenn, and Warthin-Starry, with 23 of the cases stained by Brown-Hopps. Of the stains tested, the Wright-Giemsa was the easiest to perform. The organisms on the Wright-Giemsa showed a smooth, uniform purple color, whereas the Warthin-Starry gave the organism a granular appearance that at times could be confused for silver precipitate. Both the Wright-Giemsa and Brown-Hopps stain had the highest degree of identification of the organism (defined by percent positivity). The routine use of the Wright-Giemsa stain for identification of C. pylori in antral biopsies is recommended.     

Evaluation of diagnostic methods for Helicobacter pylori gastritis

The authors evaluated the use of direct Gram-stained smears, 1- and 24-hour urease broth tests, histologic examination, and culture to detect Helicobacter pylori in 100 gastric biopsy specimens from 97 patients with epigastric symptoms. Twenty-six patients had positive cultures and 27 had H. pylori identifiable in hematoxylin and eosin-stained sections. The gastric biopsy specimens from the 29 patients with culture and/or histologic findings positive for H. pylori showed active gastritis in 27 cases (93%), compared with 26 cases (37%) without H. pylori (P less than 0.0001). Chronic gastritis was present in 25 cases (86%) with H. pylori and 40 cases (56%) without H. pylori (P less than 0.01). Twenty patients had positive Gram-stained smears. Fifteen patients had positive 1-hour urease tests, and 3 had delayed positive 24-hour urease tests. There were no false-positive Gram's stain results, three false-positive 24-hour urease tests, two false-negative histologic results, and three false-negative cultures (one inadvertently incubated anaerobically). The sensitivities of the methods were as follows: 62% for the 24-hour urease test, 69% for direct Gram's stain, 90% for culture, and 93% for histologic examination. The authors conclude that the urease test used in this study has low sensitivity and limited specificity; that the direct Gram-stained smear is a useful, highly specific, rapid screening test; and that the lengthier methods of culture and histologic examination have comparably high sensitivity for the definitive diagnosis of H. pylori gastritis.     

Transport conditions and number of biopsies necessary for culture ofHelicobacter pylori

Stuart's transport medium with a charcoal impregnated swab was tested for transport of biopsies from the gastric antrum for culture ofHelicobacter pylori. Biopsies were cultured under microaerophilic conditions either within 2 h or after a delay of 24 h at 4 °C. In 65 patients referred for gastroscopy two biopsies were taken.Helicobacter pylori was found in 39 patients. The rate of survival ofHelicobacter pylori was found to be as high in Stuart's transport medium after 24 h at 4 °C as in the paired biopsy which was cultured immediately after transportation in normal saline. In five (13 %) of 39 patients positive forHelicobacter pylori, the organism was cultured from only one of the biopsies. It is therefore recommended that two biopsies be taken for culture.     

Evaluation of a commercial latex test for serological diagnosis ofHelicobacter pylori infection in treated and untreated patients

The value of a commercially available latex test (Pyloriset) for the diagnosis ofHelicobacter pylori infection by demonstration of specific antibodies was compared with that of direct diagnostic methods such as culture, biopsy-urease test and microscopy of fuchsin-stained smears. The sera were from 136 patients who prior to this study either had or had not been treated forHelicobacter pylori-infection simultaneously with amoxicillin (3 × 750 mg/day) and metronidazole (3 × 500 mg/day) for 12 days. On average, the sensitivity of the test was 90 %. The specificity with sera from untreated patients was 75.9 %; with sera from treated patients specificity was 22.2 %, 28 % and 20 % 1, 3 and 6 months respectively after start of treatment. Only as late as one year after the onset of chemotherapy did the specificity return to 67 %. Because of its low specificity this test does not offer any advantage over other tests in the detection ofHelicobacter pylori-infection or in monitoring the chemotherapeutic success.