![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
Previous Article | Next Article 
Clinical Microbiology Reviews, April 2007, p. 280-322, Vol. 20, No. 2
0893-8512/07/$08.00+0 doi:10.1128/CMR.00033-06
Copyright © 2007, American Society for Microbiology. All Rights Reserved.
INSERM U853, F33076 Bordeaux, France, and Université Victor Segalen Bordeaux 2, Laboratoire de Bactériologie, F33076 Bordeaux, France
SUMMARY INTRODUCTION INVASIVE TESTS Culture Biopsy specimens. (i) Specimen collection. (ii) Transport of biopsy specimens. (iii) Microscopic examination of smears. (iv) Grinding of biopsy specimens. (v) Culture on plates. (vi) Broth culture. (vii) Phenotypic identification. (viii) Question of coccoidal forms. (ix) Strain maintenance. Other culture specimens. (i) Gastric juice. (ii) Blood. (iii) Liver. Gastric culture of other helicobacters. Histopathological Diagnosis Specimen collection and transport. Preparation of histological slides. Slide examination. Other histological methods. Urease Tests Principle. Factors influencing results of the urease test. Other specimens for the urease test. Molecular Methods Biopsy specimens. (i) Standard PCR for the detection of H. pylori. (ii) Standard PCR for detection of pathogenic factors of H. pylori. (iii) Real-time PCR. (iv) PCR on fixed biopsy specimens. (v) PCR for quantification. Other specimens. (i) Gastric juice. (ii) Blood. (iii) Specimens from other sites. Gastric PCR for other helicobacters. NONINVASIVE TESTS UBT Historical aspects. Principle. Factors influencing the result and standardization of the [13C]UBT. Quantitative UBT. Other methods not using breath analysis. Particularities of the [14C]UBT. Stool Tests Culture from stools. PCR on stool samples. Detection of H. pylori antigen in stools. Detection of H. pylori Antibodies Serodiagnosis. (i) Techniques for serodiagnosis. (ii) Antigens used for serodiagnosis. (iii) Antibodies detected. (iv) Factors affecting the ELISA result. (v) Development of point-of-care tests. (vi) Immunoblot analysis and detection of CagA antibodies. Detection of H. pylori antibodies in urine. Detection of H. pylori antibodies in saliva. Detection of H. pylori in the Oral Cavity HOW TO USE THE TESTS Comparative Properties of the Different Tests Used Sensitivity and specificity. Availability of the tests. Rapidity in obtaining results. Possibility of quantitative tests. Possibility to detect pathogenic properties. Globality of the test. Cost of the test. Added value of certain tests. Application in Different Clinical Settings Before treatment. Posteradication follow-up. Diagnosis for children. Diagnosis for the elderly. Diagnosis in the context of upper gastrointestinal bleeding. Diagnosis in Developing Countries ANTIMICROBIAL SUSCEPTIBILITY TESTING Phenotypes Observed Resistance Mechanisms (i) Macrolides. (ii) Amoxicillin. (iii) Tetracyclines. (iv) Fluoroquinolones. (v) Rifampins. (vi) Nitroimidazoles. Susceptibility Testing Methods (i) Phenotypic methods. (ii) Genotypic detection of resistance. Relevance of H. pylori Resistance to Antibiotics (i) Prevalence of H. pylori resistance to antibiotics. (ii) Impact on H. pylori eradication. CONCLUSION ACKNOWLEDGMENTS REFERENCES
| SUMMARY |
|---|
 Top NextReferences |
|---|
| INTRODUCTION |
|---|
|
TopPreviousNextReferences |
|---|
It is now well accepted that the most common stomach disease, peptic ulcer disease, is an infectious disease, and all consensus conferences agree that the causative agent, H. pylori, must be treated with antibiotics (131a, 295a).
Furthermore, the possibility emerged that this bacterium could be the trigger of various malignant diseases of the stomach and it is now a model for chronic bacterial infections causing cancer. The rare gastric mucosa-associated lymphoid tissue (MALT) lymphoma is the best example for which most of Bradford Hill's criteria of causality have been fulfilled, including remission of the cancer after a successful eradication of H. pylori (579). While sufficient proof is lacking for the most common gastric cancer, i.e., gastric adenocarcinoma, numerous data have highlighted the essential role of the villain (335).
The public health importance of the discovery of H. pylori and its role in stomach diseases was recognized in 2005 by the attribution of the Nobel Prize in Physiology or Medicine to B. Marshall and R. Warren. In the history of Nobel prizes, this is only the third time that the discovery of a bacterium has been acknowledged (358).
For the correct management of peptic ulcer disease and gastric MALT lymphoma, as well as obtaining information on a wide range of diseases associated with H. pylori infection, effective diagnostic methods including susceptibility testing are mandatory.
Most of the many different techniques involved in diagnosis of H. pylori infection are performed in microbiology laboratories.
The aim of this article is to review the current status of these methods and their application, highlighting the important progress which has been made in the past decade. The traditional division between invasive and noninvasive techniques will be followed.
| INVASIVE TESTS |
|---|
|
TopPreviousNextReferences |
|---|
The stomach is usually accessed by fiber optic endoscopy, and biopsy specimens are obtained. Unfortunately, with standard technology the endoscopic features of H. pylori infection are not specific. Erythema and edema may be seen, but most often, no change is observed (416). In some cases, follicular gastritis may be observed, especially in children and young adults (474), as well as major lesions, e.g., ulcers, polyps, tumors. However, recent progress has been made to magnify the gastric mucosal abnormalities with narrow band imaging (260), endocytoscopy (231), and confocal laser endomicroscopy (257). This last technique enabled detection of H. pylori by surface microscopy imaging of living tissue during ongoing endoscopy for the first time. Using two contrast stains, topical acriflavine and intravenous fluorescein, with an endomicroscope (Pentax, Tokyo, Japan), endoscopists were able to see clusters of bacteria as well as single bacterial cells stained by acriflavine both on the surface and in the deeper layer of the gastric epithelium. This report is a real breakthrough in current diagnostic possibilities (257).
To avoid endoscopy, other less invasive paths to the stomach have been proposed. It is possible to obtain gastric juice using a nasogastric tube. Gastric juice allows the detection of H. pylori by culture, staining, urease test, and PCR, but it is less reliable than gastric biopsy specimens. The string test can also be used to obtain gastric mucus (443); however, the most attractive method seems to be an extendable oro-gastric brush contained in a plastic tube (Baylor Brush, US Endoscopy, TX). The brush is swallowed, extended into the stomach to brush the mucosa three or four times, retracted in the protective sleeve, and withdrawn from the patient. This method is rapid and appears to be reliable for H. pylori infection diagnosis (187).
The potential contamination of biopsy specimens via endoscopes was a key issue in the past but now seems to have been resolved, at least in developed countries. Indeed, due to the discovery of prions and viruses, i.e., human immunodeficiency virus and hepatitis C virus (HCV), potentially transmitted by endoscopes, considerable attention has been given to the cleaning and disinfecting of endoscopes. Medical societies have even issued recommendations to use disposable forceps in some cases. A consequence for microbiologists is the elimination of gastric biopsy contaminants, e.g., Pseudomonas species and other environmental bacteria and sometimes H. pylori itself, which can be transferred from one patient to another. Contamination cases have been documented or suspected in the past, based on postendoscopy acute achlorhydria (534). Contamination of the stomach with oropharyngeal flora, and with bacteria from the intestine in the case of bacterial overgrowth, obviously remains. Pretreatment of the biopsy specimens by washing in saline could improve the recovery of H. pylori (239).
The number of biopsies which is necessary to diagnose H. pylori infection is a subject of controversy. A single biopsy specimen taken from the antrum (2 cm from the pylorus) gives good sensitivity but is not sufficient for a reliable diagnosis. Indeed, H. pylori may have a patchy distribution, and the more biopsy specimens analyzed, the higher the chance of organism detection (27). The recommendation is, therefore, to take two biopsy specimens from the antrum as well as two specimens each from the anterior and posterior corpus. There are some rare cases where the infection lies only in the corpus, but usually, H. pylori is present in all sites. After consumption of antisecretory drugs, the corpus may be the only site which remains positive.
Biopsy specimens for culture must be taken before specimens for histological examination, the latter being introduced in a fixative, otherwise there is a risk of transferring small amounts of fixative to the container for biopsy specimens to be used for culture.
(ii) Transport of biopsy specimens. A key point is the transport of the biopsy specimens from the endoscopic suite to the laboratory. Problems at this stage are surely the cause of many culture failures and have led to the negative opinion which some gastroenterologists have of culture. H. pylori is a fragile organism. It must be protected from desiccation and contact with oxygen and room temperature. It is mandatory not to expose the biopsy specimens to air and to place them either in a saline solution for short-term transport (4 h maximum) (368) or in a transport medium, usually consisting of semisolid agar, maintained at 4°C. A commercially available medium, Portagerm pylori (bioMérieux, Marcy l'Etoile, France), is effective for this purpose (192, 214). Culture may be delayed by 24 h if such transport media are used, allowing biopsies to be sent by mail in a cool transport container (Sarstedt, Nümbrecht, Germany) (192, 214). If these transport conditions cannot be used, it is better to freeze the biopsy specimens at –70°C or in liquid nitrogen in a dry tube and transport them frozen to the laboratory. Storage at 4°C in a medium containing 20% glycerol also led to H. pylori recovery in 81% of the biopsy specimens tested (203).
The proposal of direct biopsy specimen plating in the endoscopy suite is complicated, since biopsy specimen grinding and a special atmosphere for plate transportation are needed. This procedure is not used.
(iii) Microscopic examination of smears. A standard bacteriology test includes a microscopic examination of a smear prepared from the biopsy specimen or touch cytology (95). Indeed, gastric brushings (43, 59, 77) followed by examination with phase-contrast microscopy in the endoscopy suite allow a quick diagnosis (452) and are an alternative to the urease test (93) (Fig. 1).
|
Besides its ability to provide a rapid result, this method allows visualization of inflammatory cells and observation of Helicobacter heilmannii, which can be identified by its typical morphology including 6 to 8 spirals (96).
(iv) Grinding of biopsy specimens. In most instances, the bacteria are not distributed homogeneously in the biopsy specimens. Therefore, if they are only streaked on plates, several contiguous organisms will lead to one colony, whereas if the same bacteria were dispersed, several colonies would appear.
Comparison of culture performed with or without grinding showed a higher number of colonies after grinding, although the number of positive specimens did not change (182). For this reason, grinding of the biopsy specimen is mandatory. The recommendation is to use an electrical/mechanical grinder and a small volume of broth. Care must be taken to thoroughly wash and sterilize the probes. Another possibility is to perform manual grinding with disposable material. This solution avoids the risk of possible DNA contamination when performing molecular techniques.
(v) Culture on plates. Due to (i) the difficulty of growing H. pylori in broth and (ii) the usual abundance of bacteria, no enrichment has been proposed, only a direct plating is performed.
The media components include an agar base, growth supplements, and selective supplements. Most agar bases are satisfactory for growing H. pylori, e.g., brain heart agar, Columbia agar, and Wilkins Chalgren agar. Concerning the growth supplement, it is mandatory to add blood or serum, which includes numerous nutrients (vitamins and oligoelements, etc.) which enhance H. pylori growth. The proportion of blood or serum can be 5, 7, or preferably, 10%. Red blood cells can be lysed for these growth substances to be more readily available. Animal blood, e.g., sheep and horse blood, can be added, but human blood, to be used after appropriate testing, seems to confer a slight advantage (570). Other growth supplements such as egg yolk (569), charcoal (177), starch (46), bovine serum albumin, and catalase (208) are used. Cyclodextrins, which are cyclic oligosaccharides produced from starch by enzymatic treatment retaining the same properties as starch, are employed (423). More recently, ferrous sulfate, sodium pyruvate, and swine mucin have been proposed to enhance H. pylori growth (235).
Cellini et al. proposed a blood-free medium supplemented with isovitalex (2%) and hemin (10 mg/liter). They also added urea (20 g/liter) and a pH indicator (phenol red) to identify the urease-positive colonies (64). However, our experience does not favor the incorporation of a high urea concentration because the bacteria can be inhibited by the high pH induced. Indeed, pH is an important variable to consider; H. pylori grows best at a slightly acidic pH (5 to 6), in agreement with its ecological niche, the mucus layer, where a pH gradient exists (243). Another supplement which may be helpful to readily identify H. pylori colonies is 2,3,5-triphenyltetrazolium chloride (40 mg/liter). This compound is reduced by H. pylori to insoluble red formazan complexes, resulting in easily distinguished pigmented golden colonies (458).
Selective supplements are also quite important because of the presence of contaminating flora as mentioned before: buccal flora consisting mainly of gram-positive cocci, intestinal flora in the case of duodenal reflux, and bacterial overgrowth. Furthermore, Candida species may colonize ulcer craters.
Different selective supplements containing antimicrobial compounds have been proposed: vancomycin or teicoplanin to inhibit gram-positive cocci; polymyxin, nalidixic acid, colistin, trimethoprim, or cefsulodin to inhibit gram-negative rods; and nystatin or amphotericin B to inhibit fungi. The Dent supplement, a modification of Skirrow's formula in which cefsulodin replaces polymyxin and amphotericin B is added, is commercially available (105).
Several studies performed in the early days of H. pylori detection showed the importance of using both a nonselective medium and a selective medium (273, 446) or even two different selective media (13, 514).
A critical point is to use fresh media (less than a week old) which is kept in closed boxes at 4°C to maintain humidity and avoid light exposure.
Helicobacters are microaerophilic and capnophilic bacteria. When their growth was studied according to the oxygen concentration, optimal growth occurred with a pO2 of 2 to 10 KPa, whereas no growth occurred at a pO2 of air (243). In contrast to these findings, Xia et al. claimed that most of their H. pylori strains grew under aerobic conditions but with reduced cell counts and smaller colonies (585). This discrepancy can be explained by the fact that, at low bacterial concentrations, H. pylori is microaerophilic, while at high bacterial concentrations, it can grow aerobically (50). Several systems can be used to achieve a microaerobic atmosphere, from the most sophisticated systems, such as a microaerobic cabinet or an incubator with an adjustable gas level, to jars in which the adequate atmosphere is created with an automatic apparatus (Anoxomat, MART Microbiology BV, Lichtenwoorde, The Netherlands) or with H2-CO2-generating packs. Besides their cost, the gas-generating packs take several hours to produce the optimal atmosphere. The atmosphere in jars will vary according to the quantity of bacteria consuming oxygen; therefore, the gas pack should be changed every other day. While H. pylori growth is possible in a candle jar (136), it takes a longer time and results in small colonies; therefore, we do not recommend this solution.
The optimal culture temperature is 37°C, testifying to the adaptation of this bacterium to humans. Certain strains can, nevertheless, grow at 30°C or 42°C (359). For primary culture under optimal conditions, colonies may appear after 3 days and are at their optimum on day 4. However, in the case of negative culture, a 7- to 10-day incubation is recommended to ensure that the result is negative; if only a few organisms are present, this time lapse may be necessary to visualize the colonies (544).
In contrast, subcultures only take 2 to 3 days. When few colonies are present, the recommendation is to subculture by plating the colonies on a small area of the agar plate. It is important to remember that once H. pylori reaches its growth plateau, it becomes coccoidal and loses its viability, most likely due to a lack of adequate nutriments.
Culture from biopsy specimens has the potential of leading to a high sensitivity, given that only one bacterium can multiply and provide billions of bacteria. However, both strict transport conditions and careful handling in the laboratory are necessary. Culture is extremely dependent on desiccation, contact with air, and temperature, as previously stated, and sensitivity reaches 95% at best. In contrast, its specificity is 100% because, once the bacterium has grown, all phenotypic and genotypic identification tests can be performed.
Quantitative culture of viable bacteria present in gastric mucosa can be performed. However, given the variability in the biopsy specimen size (1 to 10 mg), it is recommended to weigh the biopsy specimen. The bacterial load can be extremely variable, ranging from 5 to 105 CFU/mg, in one study (24). Reproducibility is higher in the antrum versus the corpus. A high H. pylori density has been associated with bacterial virulence determinants (cagA, vacA, s1m1), gastric inflammation, and duodenal ulceration, suggesting a central role in pathogenesis (21). Given that quantitative culture is expensive, time-consuming, and not very precise, a semiquantitative evaluation is most often performed.
(vi) Broth culture. Similar media, supplements, and conditions used for agar have been applied to broth culture (492), while brain heart infusion may be preferable for studies on physiology and metabolism (315). A chemically defined broth medium consisting of serum-free Ham's F-12 medium has also been described previously (515). The growth atmosphere is important, and two recommendations have been proposed in this respect: (i) constantly check the tubes in a microaerobic atmosphere (393) or (ii) prepare a biphasic medium with a thin broth layer in a cell culture flask incubated in a microaerobic atmosphere (491). Resulting bacteria are more likely to have a typical morphology and be motile. They can also aggregate. The use of gas-permeable cell culture flasks may improve H. pylori growth (490).
To obtain an important bacterial cell mass, subculture in successively increasing volumes from 5 ml to 500 ml is recommended (492). An alternative is to use fermenters.
(vii) Phenotypic identification. The growth of small, circular, smooth colonies observed after 3 to 4 days on the selective media plated with gastric biopsy specimens is an important criterion for H. pylori identification. No hemolytic activity is readily observed but may appear after a few days at 4°C.
Microscopic examination of the cultured bacteria may show a morphology different from the bacteria present in the biopsy specimen, i.e., bacilli which are neither spiral shaped nor motile, but straight or curved. Indeed, spiral shape and motility are directly related to the viscosity of the medium (207).
The identification of cultured bacteria consists essentially of testing for the presence of certain enzymes: cytochrome oxidase, catalase, and urease (312) and eventually 
-glutamyl transpeptidase, leucine aminopeptidase, and alkaline phosphatase. Aminopeptidases and esterases (C4 to C12) are also present (359).
The family Helicobacteraceae is comprised of the genera Helicobacter and Wolinella. Helicobacteraceae and Campylobacteraceae are in the Epsilonproteobacteria, class nov. according to the latest edition of Bergey's Manual (157).
Cytochrome oxidase is present in all members of the Epsilonproteobacteria. It is usually detected with special reagents on a disk or a strip. Catalase is also present in all helicobacters and most members of the Campylobacteraceae and is detected by introducing a loopful of bacteria into a drop of hydrogen peroxide and observing a very abundant production of bubbles. Nevertheless, catalase-negative mutants of H. pylori have been described previously (571). Urease is definitely the most important enzyme for identification. To survive in its particular ecological niche, H. pylori produces large amounts of this enzyme to buffer the acidic medium and creates a microenvironment (342). Mobley et al. reported that in H. pylori, 6% of the total protein content was urease (381). When a loopful of H. pylori is put in contact with a few drops of urease medium, the color change occurs instantaneously regardless of the formulation. Other diagnostic tests are indeed either strictly based on urease, like the rapid urease test and urea breath test, or partially based, like serology and PCR, which may target urease genes.
The ApiCampy strip (bioMerieux, Marcy l'Etoile, France) allows identification of H. pylori via positive urease, -glutamyl transpeptidase, and alkaline phosphatase and negative nitrate reductase and hippuricase, all located in the first part of the strip, which explores preformed enzymes. The second part of the strip, which explores organic compounds as the only carbon source, cannot be used because growth of the organism cannot be supported by the minimal medium used.
Among tolerance tests utilized to differentiate Campylobacter species, H. pylori can grow with 2,3,5-triphenyltetrazolium chloride (0.4 and 1 mg/liter), sodium selenite (0.1%), and glycine (1%), but not with glucose (8%) and sodium chloride (3.5%). H. pylori is susceptible to cephalothin but resistant to nalidixic acid, with some exceptions (359).
When the bacteria are isolated from gastric biopsy specimens, phenotypic tests are sufficient for a precise identification. However, this is not the case when the bacteria are isolated from other specimens, e.g., stools, saliva, and the environment. Because other known or unknown bacteria sharing the same characteristics may be present, it is mandatory to confirm the identification with molecular methods. If PCR is used, several reactions targeting at least two different genes must be positive simultaneously. Sequencing of the urease gene may be worth performing because data are available for a number of Helicobacter species. Another spiral bacterium which we and others (478) identified in the stomach by culture is the urease-negative Campylobacter jejuni subsp. doylei. It was isolated for the first time from the stomach in 1985 (247).
Culture is an important step to perform further antimicrobial susceptibility testing (see below) and strain typing when necessary (49).
(viii) Question of coccoidal forms. As previously mentioned, when nutrients are lacking, H. pylori loses its spiral shape and becomes progressively coccoidal. Most people believe that these forms are both nonculturable and nonviable (278). However, others claim that some of them may be viable but nonculturable and constitute a resistant form of the bacterium (30).
The diagnostic implications are as follows: (i) a discrepancy is possible between the results from culture and from molecular tests as long as the DNA is not degraded and (ii) it is therefore important to subculture the colonies as soon as they reach their optimal size because afterwards they may die.
(ix) Strain maintenance. H. pylori is difficult to maintain. Colonies can survive on plates for a week provided they are kept in a microaerobic atmosphere at 4°C. For a long-term conservation, bacteria must be frozen at a low temperature (–70°C freezer or liquid nitrogen). Different broth media have been used, always with a cryoprotective agent, such as glycerol, either in cryotubes or on beads. The freezing-thawing process is always deadly for the bacteria, and only a small proportion survives. For this reason, it is mandatory to use bacteria in their exponential growth phase, as they are more likely to survive. Frozen H. pylori specimens can be maintained for decades at –70°C. Freezing at –20°C is insufficient, and lyophilization appears to be difficult. Loss of viability is noted particularly during the dehydration phase (428). Storage of lyophilized vials at 4°C could help bacterial survival (506).
Other culture specimens. (i) Gastric juice. Gastric juice reflects the whole stomach and is easier to obtain than biopsy specimens but still necessitates an invasive procedure, i.e., introduction of a nasogastric tube. H. pylori is found in gastric juice due to the turnover of gastric mucosa, but the sensitivity is much lower than culture from biopsy specimens, with the best results for the former being in the range of 60% (149, 555). It does not seem crucial to buffer the collected liquid.
A less invasive procedure has been proposed by Perez-Trallero et al. involving Enterotest (HDC, San José, CA), also named the string test (443). A gelatin capsule fixed to a nylon fiber is ingested by the patient. After 1 h, the fiber is removed, with the capsule having been digested. The distal part of the fiber can be used for H. pylori detection by culture but also by urease test or PCR. This technique was applied by others but with sensitivity rates ranging from 38% to 97% (304, 482, 524). Leodolter et al. proposed to use the string test after the urea breath test as a diagnostic package for the identification of treatment failure and antibiotic resistance without the necessity of upper gastrointestinal endoscopy. The sensitivity of culture from the string test was 87% compared to culture from gastric biopsy specimens (303).
To study the possible transmission of H. pylori during vomiting, Parsonnet et al. obtained vomitus after induction of emesis in 16 volunteer subjects. They were able to grow H. pylori from all of the samples, often in high quantities (434). In contrast, spontaneous vomitus allowed the isolation of H. pylori in only one of four positive cases (306). The delay before performing the analysis on unbuffered gastric juice is critical: the survival rate was 62% after 2 h versus 15% after 24 h (154).
Gastric juice is well adapted for performance of quantitative culture and estimation of the potential for transmission of this organism (153, 593). In the study of induced vomitus from infected subjects, H. pylori grew in high quantities (>103 CFU/ml), while low quantities (50 to 500 CFU/ml) were found in postemesis saliva and cathartic stools (2,000 to 5,000 CFU/ml) in most of the subjects (434).
(ii) Blood. H. pylori may reach the blood flow from the gastric mucosa in the case of ulceration and hemorrhage. However, only one isolate was found in a blood culture from a patient with a gastric lymphoma. The strain grew after 5 days in a Bactec NR730 aerobic bottle (403). Indeed, regular blood culture systems may not be optimal for H. pylori growth. Conversely, blood culture systems used for fastidious organisms, e.g., brucella broth, led to satisfactory growth when experimentally inoculated (255).
(iii) Liver. Numerous attempts have been made to isolate H. pylori or other helicobacters in the liver and bile ducts in different hepatic diseases. Only one strain has been grown from a patient with Wilson disease in Brazil (103).
Gastric culture of other helicobacters. The other Helicobacter species possibly present in the human stomach is Helicobacter heilmannii, previously known as Gastrospirillum hominis. This bacterium is commonly found in the stomachs of pets (dogs and cats) and pigs. It rarely colonizes humans and is considered a zoonosis (365).
This organism is nonculturable on standard media; it can essentially be cultured in vivo by introducing it into the mouse stomach (295). We succeeded in this endeavor three times in a clinical context.
In 1996, Andersen et al. were the first to grow H. heilmannii from a 23-year-old patient with dyspepsia on the same plate medium used in their laboratory to grow H. pylori, and growth occurred within 5 days in a standard microaerobic atmosphere. From primary culture, only 10 to 15% of the bacteria exhibited a typical morphology contrary to the majority after subculture (10). This is indeed the only positive culture on a plate for this bacterium which was later identified as Helicobacter bizzozeronii (233).
In summary, culture is still a valuable tool. It is considered the reference method when comparing the accuracy of noninvasive techniques. Its specificity is maximal, since the availability of colonies allows identification with all techniques, including sequencing of key genes (urease, vacA cytotoxin, cag pathogenicity island [PAI] open reading frames). Its sensitivity may be inferior, due to the stringent transport conditions required. Theoretically, its sensitivity should be the best because the presence of one organism in the inoculum is enough to give a positive result. However, in studies evaluating noninvasive tests, it is now the rule to consider a specimen negative by culture as H. pylori positive if both histology and urease tests are positive. Furthermore, culture is still the ideal method for antimicrobial susceptibility testing and typing, although molecular methods can also be used now directly on specimens for this purpose.
Specimen collection and transport. The usual recommendation derived from the Sydney system (109, 453) is to obtain 2 biopsy specimens from the antrum and 2 specimens from the corpus. While the rationale is essentially to type the gastritis which may be present, H. pylori detection benefits from the high number of specimens tested. As indicated before, sensitivity increases with the number of biopsies (27). Bacteria are usually present at both sites even if the lesions occur essentially in the antrum. When topographic studies of H. pylori distribution and gastritis were performed, the best site suitable for diagnosis was the lesser curvature of the midantrum, while for the corpus, there was a discrepancy between greater and lesser curvatures (159, 375, 484).
Biopsy specimens are immediately introduced into a fixative of 10% formaldehyde. Consequently, there is no transport problem, since the material can be sent at room temperature by regular mail and is no longer infectious. This fixative maintains the morphology of the bacteria, and most stains can be used. However, storage in formaldehyde is limited because, after a week, the diagnosis becomes difficult (142). The Bouin fixative must be avoided because it alters the bacterial morphology.
Preparation of histological slides. Ideally, an orientation of the biopsy specimens should be made before paraffin embedding to have sections which show the surface epithelium where bacteria are essentially located. The use of filter paper before fixation must be discouraged because the absorption of gastric mucus by the filter paper decreases sensitivity (590). It is usually necessary to cut three thin sections (3 to 5 µm) at different levels. The accuracy of this technique is very much dependent on the quality of the histological preparation, which must allow observation of the surface epithelium and the crypts.
There is no stain specific for H. pylori. The routine hematoxylin-eosin stain is not well suited for H. pylori detection because of the weak contrast between the bacteria and the mucus. There are several special stains (Table 1) which allow better visualization than the standard hematoxylin-eosin stain. This is essentially due to the acidophile characteristic of these stains which allows the staining of H. pylori DNA and not of the mucus where the bacteria are present, leading to a good contrast. The conclusion of the presence of H. pylori can be made easily, since it is seldom that bacteria other than H. pylori with similar morphology occupy this ecological niche.
|
Slide examination. The typical morphology of H. pylori in biopsy specimens is a comma or S-shaped bacillus (2.5 to 4 µm long and 0.5 to 1 µm thick). The bacteria are observed on the epithelial surface at high magnification. They adhere to the gastric mucus cells or are free in the mucus and may eventually be present in the intercellular spaces; very few bacteria, if any, are intracellular. Some organisms may be found in the canaliculi of the fundic parietal cells. Others may be present on metaplasic antral cells outside of the stomach, especially in the duodenal bulb (584) or in the esophagus. They do not colonize the intestinal metaplasia but can adhere to areas of incomplete intestinal metaplasia which consists of a hybrid epithelium with both gastric and intestinal features (426).
There are limits to this technique. First, it is necessary to obtain good quality biopsy specimens, which is not always possible, and from a series of patients, it is common to have biopsy specimens where few epithelial surfaces can be observed. Second, when there is a low number of bacteria present, and in the event that they do not have a typical morphology, it is difficult to draw a conclusion.
The same conditions as previously described for culture (i.e., unsuccessful eradication treatment, consumption of PPI or antibiotics) may also jeopardize the accuracy of histological diagnosis (237). These conditions may trigger the evolution of the bacteria from a typical morphology to a coccoidal form which does not allow a diagnosis to be made. A high proportion of coccoidal forms was noted after gastrectomy in one study (65).
The proportion of cases with a chronic active gastritis but no bacteria present on the slides is currently increasing. The main reason is probably an uncontrolled usage of PPIs. Indeed, omeprazole administered as a monotherapy for 4 weeks decreases the bacterial density in both the antrum and the corpus (185) but at a differing magnitudes from one patient to another. This effect progressively disappears a few weeks after stopping the PPIs. Similar phenomena can be observed with H2 antagonists but to a limited extent. Antibiotics used as a monotherapy also have a dramatic effect on the bacterial density even if they rarely lead to H. pylori eradication. Another entity, namely focally enhanced gastritis, seems to offer a picture similar to H. pylori gastritis. This gastritis is frequently observed in patients with Crohn's disease and, according to Pusztaszeri and Genta, is an indication to look for this disease (456).
Finally, the histological diagnosis of H. pylori infection is very much dependent on the expertise of the pathologists and on the time devoted to the diagnosis. A comparison of results obtained from experienced pathologists performing routine diagnosis highlights this problem (328, 382). The interobserver reproducibility was tested in 3 series (11, 74, 483) and measured by the Kappa test. A value of >0.75 corresponds to an excellent reproducibility, values of 0.4 to 0.75 correspond to an acceptable reproducibility, and a value of <0.4 is insufficient. The Kappa value was 0.69 and 0.74 in 2 series (11, 483), while in the third it varied from 0.39 to 0.82 according to observer pairs (74). In this last study, the intraobserver value was 0.65 to 0.88. These data cast doubt on the reliability of this diagnosis. Other studies have confirmed this problem (125, 238).
When doubt occurs, pathologists may use the presence of active gastritis as a surrogate, since this condition is almost pathognomonic of H. pylori infection. Immunostaining with a specific H. pylori antibody (Dako, Glostrup, Denmark) can also be used. In all of the studies performed, immunohistochemistry had the highest sensitivity and specificity. It allows a better interobserver agreement than routine histology and can also be performed with an autoimmunostainer (17, 110, 238, 526). In situ hybridization has been developed using a 16S rRNA gene probe labeled with digoxigenin (246) or with fluorescein (fluorescence in situ hybridization [FISH]). The latter method is now commercially available and will be described in detail later, since it can also be used for antimicrobial susceptibility testing.
Histological results must be reported according to guidelines drawn up in 1990, known as the Sydney system (453). The presence of bacteria in the corpus and in the antrum is expressed semiquantitatively on a scale of 0 to 3. In addition, the histological characteristics of the gastric mucosa (inflammation, activity, atrophy, intestinal metaplasia) are also reported. The recent update of the Sydney system proposes inclusion of biopsy specimens from the incisura, an area where premalignant lesions are commonly found (109).
Furthermore, histology allows the diagnosis of the nonculturable H. heilmannii, which is seldomly present and easily identified by its typical morphology: it is longer than H. pylori with 6 to 8 spirals, nonadherent to epithelial cells, and commonly localized in aggregates in the crypts' lumen (Fig. 2). It also induces gastritis, which is sometimes transient, and may lead to gastric ulcer and gastric MALT lymphoma (96, 216, 394, 505). The FISH method can also be used to detect this bacterium (528).
|
Other histological methods. Electron microscopy can be used to detect H. pylori on the mucosa (145). However, this technique is too time-consuming and difficult to employ routinely.
Biopsy specimens frozen with a cryostat before an extemporaneous examination have been used as a rapid method, but their interpretation is difficult (244).
Principle. The discovery that H. pylori was a strong urease producer was made by Langenberg et al. early in the bacterium's history (290). It was first applied to direct rapid diagnosis by McNulty and Wise (356).
To adapt to its special ecological niche where the concentration of urea diffusing from blood to the gastric mucosa is low, H. pylori produces large amounts of urease. H. pylori urease also has the highest specific activity (36 ± 28 µmol/min/mg of protein) among bacterial ureases (381).
The other urease-positive bacteria present in the gastric mucosa, i.e., streptococci and staphylococci, produce a lower amount of urease, which does not interfere in a short-time detection (<2 h), rendering the method specific to H. pylori.
When a biopsy specimen containing H. pylori is introduced into a urea-rich medium, the urease breaks the urea down into carbon dioxide and ammonia. The ammonium ion increases the pH, and a pH indicator, e.g., phenol red, changes color, in this case from yellow to red or violet.
The different urease media commonly utilized in bacteriology, i.e., Christensen medium and urea-indole medium, can be used, but specific media are preferable.
Different ways to improve the sensitivity of the tests include: (i) increasing the urea concentration from 2 to 10% (41), (ii) incubating at a higher temperature (37°C) (287, 595), and (iii) suppression of the buffer (15), while the size of the biopsy does not appear to be important (286, 594). Unbuffered media have a limited shelf life (<5 days at 4°C) (406).
The first-generation commercial kits were agar based, e.g., the CLO test (Kimberley-Clark, Neenah, WI). The new generation kits are strip-based tests with two areas separated by a microporous membrane, one where the urease hydrolyzes urea and the other where NH3 is trapped and changes the pH (PyloriTek, Serim, Elkhart, IN).
The agar-based tests exhibit a good sensitivity only after 24 h (90 to 95%), compared to 70 to 80% after 1 h. Therefore, they cannot be classified as rapid tests after such a long delay. Furthermore, other urease-positive bacteria from the mouth can decrease the specificity when reading is performed after 24 h (489).
H. heilmannii, which may be present in the stomach, as previously indicated, is also a urease-positive bacterium but may not give a positive result, as the bacterial load is often limited. Therefore, the distinction between H. heilmannii and H. pylori is made by histological examination as previously indicated.
Attempts to obtain a rapid quantification of H. pylori based on the urease produced have been performed using an ammonia electrode measuring the ammonia liberated from samples placed in a urea solution (51).
Factors influencing results of the urease test. A limit to the urease test is the bacterial load necessary to obtain a sufficient sensitivity. A semiquantitative evaluation of the bacteria present by histology clearly showed that false-negative urease tests corresponded to the lowest histological scores for H. pylori (38, 522, 597). It seems that at least 105 bacteria are necessary for a valid result. This amount may not be present 4 weeks after the failure of eradication therapy, which makes this test less advisable for posteradication follow-up. Treatment with PPI may also jeopardize the result. By changing the milieu where the bacteria are present, especially the antrum, PPI renders it inhospitable and the bacterial load decreases. In addition, PPI themselves may have antiurease properties (533). Another reason for a false-negative test is the presence of intestinal metaplasia, which also corresponds to an inhospitable environment for H. pylori.
Per endoscopy, stains such as methylene blue may negate the urease test (198) so the specimens for this test must be taken before employing these stains.
The great progress in this area was the introduction of a strip-based test (PyloriTek) in 1995. In the first study, Rogge et al. compared this new test to the CLO test using histological detection as the reference. PyloriTek showed a 99% sensitivity and 95% specificity after 2 h, which is superior to those of the CLO test (468).
These results have been confirmed since then in several studies (288, 455, 592, 596). Indeed, the sensitivity of the final reading is not significantly different from that of the CLO test, but the last reading can be done after 1 h versus 24 h for the CLO test. For routine use, most endoscopists read the CLO test earlier than recommended, which leads to a marked decrease (20%) in sensitivity (288, 454).
Other specimens for the urease test. The use of the urease test on specimens other than gastric biopsy specimens, e.g., oral specimens, must definitely be discouraged because many other urease-positive bacteria (Staphylococcus spp., Streptococcus spp., etc.) can be present and give false-positive results.
Biopsy specimens. (i) Standard PCR for the detection of H. pylori. The choice of a target is important in designing the primers which must be specific for H. pylori but conserved in all strains of the species. It is therefore necessary to know the DNA sequence of the target in as many strains of H. pylori and as many strains of other bacterial species as possible. The first targets used were the genes of the urease operon: ureA (78) and glmM, formerly named ureC (107), or the 16S rRNA gene (222, 226) (Table 2). While highly conserved in bacteria, the 16S rRNA gene exhibits sequences which are specific for different species. However, the specificity of this PCR has been challenged because the 109-bp amplicon obtained with primers targeting the 16S rRNA gene reacted with human tissue samples (73). Other genes with unknown function, e.g., the gene encoding a 26-kDa protein identified as tsaA (201), or random sequences were used (540).
|
Theoretically, PCR can detect only one copy of the target DNA when tested in water, but the amplification efficacy is not as good when biological material is concerned. Most studies have shown that standard PCR sensitivity is similar to that of culture for pretreatment diagnosis (31, 132, 297, 386, 540, 553). Several methods have been proposed to improve sensitivity.
To improve DNA extraction, phenol extraction or the use of special extraction kits (QIAGEN, Valencia, CA) to eliminate PCR inhibitors is superior to simple boiling (563). An internal control may help (520) but does not seem to be mandatory.
The use of a nested or seminested PCR was suggested. However, the use of nested PCR increases the risk of airborne contamination by amplicons and must be discouraged (131). Performing a second PCR with the same primer has also been proposed (502).
To increase the sensitivity of amplicon detection, a probe hybridization method can be used instead of the standard detection of the amplicons based on their size by electrophoresis (123, 567). New methods in liquid phase (DNA-enzyme immunoassay) (282, 385) and the reverse dot blot line probe assay (LiPA) (552) have been proposed. In addition, this approach definitely increases the PCR's specificity.
The use of reverse transcription-PCR (RT-PCR) has been reported by several authors (123, 126, 418, 438). Given that it is based on mRNA, it determines the viability of the bacteria present, but no improvement in sensitivity has been shown.
A significant improvement in sensitivity occurred with the introduction of real-time PCR protocols. As is always the case when a new method is more sensitive than the reference method, it is necessary to prove that the results are not falsely positive. One possibility is to conduct a follow-up of the patient over a few months to confirm the reality of the infection.
Specificity is not usually a problem when dealing with gastric biopsy specimens. However, the possibility of false positives arises when endoscopes (470) or grinding apparati in the lab are not correctly cleaned. It is recommended to use strict cleaning procedures in the endoscopy suite and disposable material in the lab to avoid such contamination. An important risk of contamination occurs when the amplicons are analyzed by electrophoresis. For this reason, there are strict rules stating that three independent rooms must be used for performing the different steps of a PCR.
An advantage of PCR is that DNA does not require strict transport conditions, and it has been performed on urease tests sent by mail (108, 314, 316).
(ii) Standard PCR for detection of pathogenic factors of H. pylori. An important application of standard PCR is the detection of specific pathogenic factors of H. pylori. There are two main pathogenic factors: the cag PAI and the polymorphism of the vacA gene. Other genes involved in adherence (babA2, sabA) or in pathogenicity (oipA, dupA, iceA) can also be detected by PCR.
Numerous studies have shown that strains harboring the cag PAI are associated with more severe diseases, especially peptic ulcer disease (84, 281) and gastric adenocarcinoma (36), as well as precancerous lesions (130, 450) and extradigestive diseases (148, 448). In contrast, no association was found with gastric MALT lymphoma except for the vacA m2 allele (100, 296), and a negative association exists with gastroesophageal reflux (319). However, these associations have not been found everywhere. In fact, in Asia, most of the strains are cag PAI positive, and therefore, this pathogenic association cannot be used in case control studies (497, 501).
The main gene of the cag PAI is cagA. It was the first to be used as a marker for the presence of cag PAI (281). We now know that CagA proteins induce morphological alterations of the gastric epithelial cell, while other genes of the cag PAI lead to production of interleukin 8 (86, 588) via the intracellular Nod1 receptor and the nuclear factor 
B pathway. The chemokine production is recognized as being responsible for an increased gastric inflammation and subsequent disease development.
While detection of the cag PAI has limited consequences on an individual basis, it has interesting implications in epidemiological and pathogenic studies. In addition to the disease association already mentioned, cagA-positive strains are easier to eradicate (44, 548).
It is possible to look for antibodies directed against the CagA protein or to detect the cagA gene. The first studies were performed on H. pylori strains by PCR. The variability in cagA sequences favors the successive use of two PCRs (281) or a Southern hybridization (501) to avoid false-negative results. The same PCR can be applied directly on DNA isolated from gastric biopsy specimens and gives similar results with bacterial colonies (67, 283, 433). The former is the preferred technique because it gives a picture of what exists in vivo (194). Material from a positive CLO test as well as from formalin-fixed embedded tissue may allow cagA detection (349, 487).
The presence of both cagA-positive and cagA-negative isolates in the same patient has been described previously (94, 137, 523). Comparison of the isolates by molecular typing showed that they were indeed the same strain (523), indicating the possible deletion of part of or the entire cag PAI. By using in situ hybridization with biotylinated probes for an H. pylori common gene and an oligonucleotide specific for cagA, it was possible to identify cagA-negative bacteria in the mucous gel of the apical epithelial surface, whereas cagA-positive bacteria colonized the immediate vicinity of epithelial cells or the intercellular spaces (57).
H. pylori also produces a vacuolating cytotoxin, VacA, which has been associated with the more severe diseases, e.g., peptic ulcer disease (144) and gastric adenocarcinoma (371). The gene encoding this cytotoxin is present in all strains but exhibits a mosaicism in the terminal (s) and median (m) regions. There are several alleles corresponding to various amounts of toxin produced: s1m1 corresponds to the highest production, followed by s1m2, while strains with the s2m2 allele do not produce any toxin (19). A PCR has been proposed to identify the genotype (20).
When both cagA and vacA detection is performed, a strong association exists between the presence of cagA and vacA s1, corresponding to strains with the highest production of cytotoxin (23, 202, 217, 475, 547, 589) as well as more severe pathology and disease.
A test has been designed to detect H. pylori and its main virulence factors: the INNO-LiPA (Innogenetics, Ghent, Belgium). This assay is comprised of a strip containing multiple specific probes for the vacA s region, the vacA m region and cagA. Some of the corresponding H. pylori genes are amplified with a biotin-labeled PCR primer and subsequently analyzed by a single-step reverse hybridization on the strip (552). Other technical developments include a multiplex PCR for vacA and cagA genotypes (67) and real-time PCR melting curve analysis using SYBR green I dye (477).
(iii) Real-time PCR. The new real-time PCR technique is a breakthrough in the diagnosis of H. pylori because it allows not only a quick and precise detection of H. pylori but also its quantification and the detection of point mutations associated with antibiotic resistance.
The principle consists of following the increase in amplicons formed in real time. This can be done by using an intercalating agent, e.g., SYBR green I dye, or the fluorescence resonance energy transfer (FRET) principle (576).
Primers specific for the target gene are designed, as are a biprobe on the amplicon, the acceptor probe (sensor probe) 5' labeled with LC-Red 640, which hybridizes where the mutation site is located, and a donor probe (anchor probe) 3' labeled with fluorescein, which hybridizes 3 to 5 nucleotides upstream. When the anchor probe fluorophor is excited, it transfers the energy to the sensor probe fluorophore, which emits a signal. To detect point mutations or a polymorphism, a melting curve analysis is performed.
Since the essential application is not limited to H. pylori detection, real-time PCR will be reviewed with susceptibility testing.
(iv) PCR on fixed biopsy specimens. Performing PCR on formaldehyde-fixed paraffin-embedded material is a very attractive possibility because (i) it would allow the study of archive material and (ii) for fresh biopsies, there would be no maintenance problem. Unfortunately, fixed specimens are far inferior to frozen material. The best results are obtained with PCRs producing short amplicons because DNA can be broken by fixatives.
(v) PCR for quantification. PCR is also used for H. pylori quantitation first by using a competitive PCR, followed by the development of a real-time PCR which has considerably facilitated the process.
Competitive PCR is based on the coamplification of an internal standard (PCR mimic) and a target DNA sequence of different sizes with the same set of primers. Various amounts of PCR mimic are added to the target, and they compete for primer binding and amplification. A visual comparison allows the detection of bands with the same intensity indicating the amount of target DNA present (389). A colorimetric detection is also possible (430). The quantity of H. pylori in gastric mucus correlated with other invasive tests as well as with the urea breath test (UBT) in a study by Furuta et al. (152). They also used this method for patient posteradication follow-up with a high predictive value (153).
Real-time PCR was first used to quantify H. pylori DNA in gastric biopsy specimens in 2002. He et al. carried out a noncompetitive PCR on a LightCycler apparatus and FRET technology (209). The detection limit was 103 target copies (ureC) and there was a linear variation from 103 to 109 CFU/ml. While this study brings forth interesting data, the presence of target DNA in supposedly H. pylori-negative biopsy specimens casts a doubt on the specificity. The authors did not use surrogate markers of infection or follow up in these discrepant cases. The same method was used by Lascols et al. but with another target, 132 bp of the 23S rRNA gene. A linear correlation was also found with as few as 300 CFU/ml up to 3 x 108 CFU/ml. The correlation with other invasive tests used semiquantitatively (histology, culture) was excellent (r, 0.86). Sensitivity and specificity of the technique were 97% and 94.6%, respectively (292). Another technological tool, the TaqMan, was used on paraffin sections of the biopsy samples (264). There was a good correlation with the UBT values.
DNA extraction is an important step in these procedures. It was evaluated in a mouse stomach model. Homogenization with glass beads, followed by use of the QIAGEN DNA mini tissue kit, was found to be the most suitable method (472). The same model was used to test the reliability of real-time RT-PCR to assess the abundance of transcripts in the gastric mucosa. Comparison to infected humans confirmed the value of this tool (469).
Other specimens. (i) Gastric juice. As for culture, PCR targeting mainly ureA genes has been applied to gastric juice aspirates with a good sensitivity and specificity (254, 347, 535, 572).
Since the introduction of the string test (442), this method of gastric juice collection has been favored (112, 140, 471, 591). The cagA gene, which is highly prevalent in the Far East, has been the target in a study using gastric juice in Taiwan (564).
(ii) Blood. PCR is not performed on blood samples for the detection of H. pylori. Interestingly, Dore et al. tested serum samples from a small number of infected patients and controls with genus-specific primers (namely C97-C98), as well as with primers designed on the conserved region of the vacA gene; 65% of the H. pylori-positive patients were positive with the Helicobacter genus primers and 75% of them also resulted in the amplification of vacA sequences. Cloning and sequencing of 16S rRNA gene amplicons of 3 samples had 99% identity with H. pylori. This surprising finding indicates that H. pylori DNA might circulate in peripheral blood (116).
(iii) Specimens from other sites. PCR was used to detect H. pylori in various sites along the digestive tract. The bacterium was detected in 2 of 46 appendix specimens from patients with appendicitis (4%) (437), in 4 of 12 ethmoid specimens from patients with chronic rhinosinusitis (33%) (429), and in 7 of 30 tonsil and adenoid tissue samples (30%) (76), probably following reflux.
PCR on liver specimens also yielded positive results for H. pylori-like organisms which seem to be associated with more severe diseases (cirrhosis, hepatocellular carcinoma) versus hepatitis alone in HCV-positive patients (467). The low discriminant power of 16S rRNA gene sequencing for species identification in the genus Helicobacter limits a definitive incrimination of H. pylori (424).
Attempts to detect enterohepatic helicobacters from the intestinal mucosa of patients with inflammatory bowel disease have also allowed the detection of H. pylori, probably as a bystander rather than a pathogen. Indeed, H. pylori crosses the intestine to be shed in feces but as a nonviable organism (422, 600).
The problem of specificity of PCR becomes fundamental in specimens from sites other than the stomach. At least two PCRs based on different target genes must be positive to conclude an H. pylori infection, as indicated before (76, 467).
Gastric PCR for other helicobacters. The tightly spiral H. heilmannii organisns are comprised, in fact, of 2 types: H. heilmannii I corresponds to a bacterium from the pig (58) and is also known as "Candidatus Helicobacter suis"; H. heilmannii II corresponds to 3 species, H. bizzozeronii, Helicobacter felis, and Helicobacter salomonis, usually present in cats and dogs. A 16S rRNA gene-based PCR was developed to detect these three species as a group. A small target DNA fragment (78 bp) enabled the study of formalin-fixed paraffin-embedded material (98). A multiplex PCR based on the tRNA intergenic spacers and on the urease gene allowed a discrimination of the three species and, combined with a 16S rRNA gene PCR for "Candidatus Helicobacter suis," covers all of the H. heilmannii types (25, 99).
A novel PCR targeting 16S rRNA gene has been designed to detect H. heilmannii-like organisms which are also pathogenic and could remain undetected by microscopic observation. Screening of 131 gastric biopsy specimens from dyspeptic patients with this assay revealed a prevalence of 2.3% in Southeast England. It was then combined with an H. pylori vacA PCR in a multiplex PCR assay allowing the detection of both pathogens (70).
The use of genus-specific primers based on the 16S rRNA gene (positive) in addition to glmM primers (negative) allowed the detection of gastric infection by Helicobacter species other than H. pylori. The 2 cases detected out of 126 were identified as Helicobacter cinaedi, an enterohepatic helicobacter, after cloning and sequencing (439).
| NONINVASIVE TESTS |
|---|
|
TopPreviousNextReferences |
|---|
The major problem is inherent to the invasiveness of the procedure. Despite the existence of small endoscopes like nasogastric endoscopes, the procedure still causes discomfort, which an increasing number of patients want to avoid. Anesthesia can be used, but it increases the risk of the procedure, which albeit very low, must still be considered. The previously mentioned risk of contamination by viruses such as human immunodeficiency virus or HCV, although theoretically nonexistent, is also a threat to some patients. Moreover, the cost of endoscopy is high, with an additional cost if anesthesia is performed; there is a need for disposable forceps, and the patients lose one working day or more.
Another limit of a direct diagnostic procedure is the possibility to explore only a small part of the stomach, i.e., a few mm2 of a total surface area of 800 cm2 (85), leading to possible sampling errors and the subsequent need to do several biopsies.
Therefore, numerous attempts have been made to develop diagnostic methods which avoid endoscopy. Since H. pylori is an infectious agent, the first method used was serology. However, due to the difficulty in obtaining an optimal specificity, other methods have been proposed: UBT, stool antigen test, and most recently, detection of specific antibodies in urine or saliva.
The same principle was applied in 1987 by Graham et al. using urea labeled with 13C, a nonradioactive isotope, to detect H. pylori in humans (186). Marshall and Surveyor then described the [14C]UBT in 1988 for the same purpose (345).
Principle. A solution of labeled urea ingested by the patient is rapidly hydrolyzed by H. pylori urease if this organism is present in the stomach; the labeled CO2 is a
