Article Text

PDF

New immunological assays for the diagnosis ofHelicobacter pyloriinfection
  1. D Vairaa,
  2. J Holtonb,
  3. M Menegattia,
  4. C Riccia,
  5. F Landia,
  6. A Ali’a,
  7. L Gattaa,
  8. C Acciardia,
  9. S Farinellia,
  10. M Crosattia,
  11. S Berardia,
  12. M Migliolia
  1. aDepartment of Internal Medicine, University of Bologna, Bologna, Italy, bDepartment of Bacteriology, University College London Medical School, London, UK
  1. Dr D Vaira, Clinica Medica I, Università di Bologna, Policlinico S. Orsola, Bologna, Italy.

Statistics from Altmetric.com

Summary

There are several types of immunological tests available for the diagnosis and management of Helicobacter pylori infection. Most commercially available serological kits use the enzyme linked immunosorbent assay (ELISA) test format. Originally the kits used crude antigen preparations although many of the newer kits use a more purified antigen preparation, with often increased specificity but lower sensitivity. Near patient test kits are based either on latex agglutination or immunochromatography. Generally they have low sensitivities compared with laboratory tests. Western blotting, ELISA, and recombinant immunoblot assays (RIBA) have also been developed into commercially available kits and can be used to indicate the presence of specific virulence markers. An antigen detection kit has been developed for the detection ofHelicobacter pylori in faeces. Immunological reagents have also been combined with other diagnostic modalities to develop immunohistochemical stains and DNA immunoassays.

Helicobacter pylori is now recognised as the cause of gastritis and most cases of peptic ulcer disease (PUD); its long term carriage increases the risk of gastric adenocarcinoma sixfold and it is designated as a class I carcinogen.1 H pylori has also been implicated as a cause of gastric mucosa associated lymphoid tissue lymphomas. Its relation to non-ulcer dyspepsia remains controversial. Additionally, long term carriage of the organism may be associated with short stature in young girls and, in the general population, as a possible risk factor for the development of vasospastic disorders and possibly skin immunopathology such as urticaria. With the recognition of H pylori as an important human pathogen,2 it has become one of the growing number of organisms to have its complete genome sequence mapped.3

Serology is an important method of determining colonisation status and can be used for diagnosis, as a screening procedure, or to follow the efficacy of eradication regimens. Most serological assays are in the ELISA format although some are based on the latex agglutination reaction. These latter are used principally as near patient assays. Most assays detect IgG in serum although some detect serum IgA. More recently developed assays detect IgA in saliva and the production of affinity purified antibodies has led to the development of an antigen detection assay for faecal specimens. Serological reagents have also been used in immunocytochemistry and to speed up the detection of amplified products of the polymerase chain reaction (PCR)–DNA immunoassays.

Current non-invasive immunological tests

SEROLOGY

Several different techniques exist for antibody detection. Solid phase assays are by far the most numerous and convenient. ELISA is the most often used serological format. It is usually laboratory based, requiring both some technical competence to perform and the availability of an instrument to measure the optical density of the coloured product. Alternatively, latex agglutination tests are most suitable as near patient tests because they are technically simple to perform and provide a result within minutes rather than the hour or two for ELISA tests. In the technique for simultaneously detecting a serological response to all antigens, represented by a single value of colour intensity (optical density) as in the ELISA test, both the number and intensity of staining of antibody bound protein bands or antibodies to specific antigens are taken into account. A commercial western blot test (Helicoblot 2) for the detection ofH pylori is also available.

ELISA kits

There have been many publications comparing either single or multiple kits4-14 in defined populations. Tables 1 and 2list the kits and their reported sensitivity, specificity, and positive and negative predictive values.

Table 1

Commercially produced serological assays for detection of Helicobacter pylori infection

Table 2

Comparison of commercially available ELISA kits for detection of Helicobacter pylori infection

Latex agglutination kits

The agglutination format is not as frequently used as the ELISA format but evaluations have been published.15-21 The latex agglutination test is more convenient than the ELISA format for near patient testing and has comparable sensitivity and specificity (table 3).

Table 3

Comparison of commercially available near patient tests for detection of Helicobacter pylori infection

New immunological assays for H pylori

HISTOLOGY

The use of stains such as the Giemsa stain to detectH pylori in tissue sections relies on the experience of the histopathologist to recognise the typical morphological appearance of the organism. Clearly, small numbers of organisms may be missed. Developments of current histological methods include the use of immunohistochemistry in order to increase both the sensitivity and specificity of the method. In a comparison of immunohistochemistry, using a polyclonal antiserum, with the Giemsa and Warthin-Starry stains, the method had a higher specificity and a lower interobserver variation.22 Immunocytochemistry also performs significantly better compared with culture and biopsy urease test when small numbers of bacteria are present and following eradication treatment.23 Other histological developments include the fixing of the mucus layer using Carnoy’s solution combined with immunohistochemistry, which reduces both the false positive and negative results obtained with conventional histological stains.24

PCR AND DNA-ENZYME IMMUNOASSAY

Although most diagnostic PCR amplicons are detected by gel electrophoresis, there is a growing trend towards using a colorimetric method of detection as this is more rapid and can be automated.

Three DNA enzyme immunoassays have been developed to detectH pylori. The GEN-ETI-K DEIA (Sorin, Italy) uses streptavidin coated microwell plates to which is added a biotinylated specific probe based on the UreC gene of helicobacter. The amplified product is added to the plate and the duplex DNA detected with an enzyme linked antibody against double stranded DNA. PYLORI-PROB (Biocode, Belgium) is a solid phase sandwich hybridisation assay. A specific capture probe (also based on the UreC gene) is bound to the microwell plate. The amplified product is added to the well and detected by a biotinylated probe to which are added streptavidin conjugated peroxidase and substrate. PCR-ELISA (Boehringer, Germany) uses a capture probe labelled with biotin (UreC) bound to the solid phase in a streptavidin coated microwell plate to which is added the amplified product labelled with digoxigenin and the hybridised product is detected with anti-digoxigenin peroxidase.

Assessment of these three kits25 was performed on biopsy specimens from 199 patients and compared with PCR with gel electrophoresis and histology/culture as a gold standard. The colorimetric DNA immunoassays were 100 times more sensitive than the standard PCR and results were obtained within four hours. The sensitivity and specificity of the three tests were 85% and 88%, respectively, for GEN-ETI-K DEIA, 87% and 82% for PYLORI-PROB, and 80% and 68% for PCR-ELISA. The study assessed patients before and after treatment and before treatment the specificity of PYLORI-PROB and PCR-ELISA was only about 50%. This apparent lack of specificity may highlight limitations in the “gold standard” rather than reflect on the DNA-enzyme immunoassay kits.

SEROLOGY

There are several new areas in which serological tests are being developed to detect H pylori. These include the detection of IgA; virulence specific ELISA; immunoblot techniques; immunological tests for salivary antibodies; near patient tests; and antigen detection. Alternative formats that have been used as serological assays include flow microsphere immunofluorescence and immunochromatography, this latter for the rapid near patient testing of whole (capillary) blood. In the former, polystyrene microbeads are coated with complex antigen of H pylori derived from a whole cell sonicate or urease enriched fraction. The serum containing the analyte (antibodies toH pylori) is mixed with the coated microspheres and binding is detected with fluorescein isothiocyanate labelled antihuman IgG and a flow cytometer. This technique does not readily lend itself to commercialisation as not every laboratory has a flow cytometer. The principle of the immunochromatography format is binding of an antigen (in this case a purified antigen ofH pylori) to a membrane, addition of whole blood, and subsequent addition of anti-human IgG labelled with coloured particles. The presence of anti-H pyloriantibodies in the blood is detected by a coloured spot on the membrane. An adjacent area of the membrane with bound human IgG acts as control. Alternative arrangements of this format can be used in which the antigen coated coloured particles bound to anti-H pylori antibodies diffuse along the length of the membrane and are trapped by a line of antihuman IgG antibodies, thus concentrating the colour at a specific position.

Finally, isoelectric focusing can be used to investigate the serological response to infection, including that withH pylori, although it is more a research method rather than a commercial test used in the management of disease. The principle is the focusing of serum antibodies to their pH by means of electrophoresis along a pH gradient, followed by transfer to a membrane loaded with antigen. Bound antibody is detected in a fashion similar to that of western blotting. Using this technique one can detect a monoclonal, oligoclonal, or polyclonal response representing the cellular dynamics of antibody production.

Serum antibody assays

Assays for IgA have been neglected in favour of IgG. However, some commercial kits do assess serum IgA and in one study of 400 patients the sensitivity and specificity of the IgA assay was comparable with that of IgG.26 Whether IgA levels give any clinical advantage over IgG is currently unknown.

The detection of CagA and VacA by ELISA or immunoblotting may provide additional information which could be of clinical relevance. Infection by Cag and Vac positive strains (type I strains) is generally more likely to be associated with more serious gastroduodenal disease compared with negative (type II) strains. This association, however, is not seen in all countries. In one study of 98 patients detection of 125 kDa (Cag), 87 kDa (Vac), and a hitherto unidentified 35 kDa component of helicobacter was correlated with ulcer disease.27

Salivary antibody assay

Because of the ease of collection, particularly in children, salivary antibody assays may be particularly useful. In one study of 157 patients referred for investigation of dyspepsia, using the Helisal (Cortex) salivary kit, the positive predictive value was very low (sensitivity 85%; specificity 55%; positive predictive value 45%; negative predictive value 90%) when compared with histology in a population in which the prevalence of helicobacter infection was 30%.28 Similarly, in a study of 86 patients when compared with culture, histology, and the rapid urease test (RUT), the Helisal kit had a sensitivity and specificity of 88% and 71%, respectively, but was inferior to assessment on serum.29

Using the OraSure salivary device (Epitope Inc.) to collect salivary antibodies from 100 children, sensitivity and specificity of 100% and 61%, respectively, were obtained with positive and negative predictive values of 100% and 92% when compared with a urea breath test (UBT).30 Similarly good results were obtained in another study using the OraSure test. In this study,31 740 children were tested with the saliva collected by the OraSure device and the serology performed by HM-CAP EIA (Enteric Products), giving a sensitivity and specificity of 85% and 93%. These results are encouraging and with further development salivary antibody tests may be of diagnostic utility.

Western blotting and CagA ELISA

Because western blotting determines the serological response to a range of antigens from an organism, it could provide more information than an ELISA. Initial studies showed that serological response to certain antigens was strongly correlated with infection and in particular the 120–138 kDa CagA protein was associated with PUD and gastric cancer. In one study, comparing immunoblotting with an in-house ELISA and a commercial ELISA (Color Vue EIA) that uses an unspecified purified antigen, western blotting was more sensitive and specific than the enzyme immunoassays28 and serological reactions to specific proteins—for example, CagA could be detected. CagA is a highly immunogenic protein whose function is unknown. It is only present in about 50–60% of strains; the presence of CagA in a strain is correlated with increased interleukin 8 secretion by the gastric epithelium and is possibly associated with the presence of PUD and gastric cancer, at least in some populations. The detection of a serological response may therefore give clinically relevant information about the infecting strain. Western blotting has been developed into commercial kits, Helico-blot 2.0 and RIBA strip; the latter identifies immunological response to recombinant CagA and VacA proteins. In a large study of over 3000 subjects,32 the RIBA results were compared with 13C-UBT and western blotting and had a sensitivity and specificity of 92%. Furthermore, a positive reaction to the lysate or Cag and Vac correlated with current infection, but a positive result with either Cag or Vac alone indicated a past infection. In some studies, however, serological detection of antibodies to Cag was no better than detection of anti-helicobacter antibodies.33 In a comparison of Helicoblot 2, RIDA Blot Helicobacter, and RIBA HP SIA, the positive and negative predictive values for diagnosis of colonisation were 97% and 90% (HelicoBlot 2); 94% and 92% (RIDA), and 98% and 85% (RIBA). For CagA status, however, the positive and negative predictive values were, respectively, 96% and 55% (HelicoBlot), 94% and 100% (RIDA), and 95% and 55% (RIBA); and for VacA status, 100% and 46% (HelicoBlot), 78% and 100% (RIDA), and 95% and 76% (RIBA). Significant false positive and negative results were obtained. When used for pre-endoscopy screening the tests missed a significant number of patients with serious pathology.34

Near patient tests

There is an increasing requirement for near patient testing as more patients are managed by general practitioners on a test and treat basis. Several new near patient tests are available but generally they lack diagnostic accuracy when compared with laboratory ELISAs. Most of the tests are one step tests using whole blood but others require serum separation, which diminishes their usefulness as near patient kits. Also, with one test kit, variations in sensitivity and specificity were noted depending on whether capillary or venous blood was used. Helisal Rapid blood test had a sensitivity and specificity of 88% and 91%, and positive and negative predictive values of 92% and 86% when compared with histology, culture, RUT, and 13C-UBT in 154 dyspeptic patients.35 These results compared with 93% and 87% sensitivity and specificity for an ELISA test (Helico G) in 109 of the patients. Another assessment on 203 patients,36compared with RUT/histology, gave a sensitivity and specificity of 82% and 91%; in this study there was concurrence of results whether venous or capillary blood was used. When Helisal was compared with a laboratory ELISA its sensitivity and specificity were 83% and 78%.37 Other studies have found much lower specificity (55%) when compared with RUT/histology/culture.38 In a study of 200 patients, although the sensitivity was high (91–92%) the specificity was low (56–62%).39 In a study comparing Helisal with the UBT on 351 patients the specificity of the near patient test was poor compared with the UBT and the specificity of the kit decreased with increasing age of the patient.40

A study of the accuracy of a new rapid whole blood test kit Pyloriset Screen, compared with Pyloriset EIA and Pyloriset Dry, on 207 patients showed it had a sensitivity and specificity of 95% and 94%, respectively, with positive and negative predictive values of 91% and 97%.41 The QuickVue one step blood test had a sensitivity and specificity of 82% and 83% in a retrospective analysis of 193 patients and 89% and 93% in a prospective study on 24 patients. If venous blood was used rather than capillary blood the sensitivity decreased to 73%.42 Hp Chek was tested on 287 patients in 10 centres in the USA and gave a result within nine minutes. The results were similar if whole blood or serum was used and the sensitivity, specificity, and positive and negative predictive values were 88%, 85% 83%, and 90%, respectively. Rapid immunochromatographic tests are also available. In a study of 1209 samples from children, the sensitivity and specificity of FlexSure HP were 83–90% and 90–100%, respectively, but the positive predictive value was low in asymptomatic children compared with that of an ELISA test.43 However, a further study found FlexSure to have a positive predictive value of 88% which was equivalent to a laboratory based ELISA result.44

Interlaboratory and intralaboratory comparisons have not yet been performed and published for these kits and a recent publication does not support the use of these kits as presently formulated for near patient testing.45

Antigen detection

One of the most promising developments for the detection ofH pylori is faecal antigen detection because of its speed, technical simplicity, and ease of sample collection. The Premier platinum HpSA EIA (Meridian Diagnostics, USA) uses an affinity purified polyclonal antibody attached to microwell plates to which is added a faecal suspension followed by a peroxidase conjugated antibody and substrate. This kit has been assessed in several studies. In a European multicentre study from 12 participating centres, 501 patients were analysed and the efficacy compared with histology, RUT, culture, and UBT.46 The sensitivity and specificity were reported as 94% and 91%, respectively, with positive and negative predictive values of 93% and 92% in a population where the overall prevalence was 55% helicobacter positive. A smaller number of patients was studied four weeks after the end of eradication treatment and gave a sensitivity and specificity of 100% and 90%. The faecal antigen kit has been tested in other studies,47-49 all giving comparable sensitivities and specificities (91–98% and 83–96%). After treatment, the antigen seems to disappear quite rapidly (within a few days) from the faeces50 but the specificity was recorded as 79%.47 However, in a further study comparing HpSA with PCR, antigen and DNA were still detectable in the faeces 35 days after treatment.51 Further development of the test is likely to improve its utility in monitoring eradication treatment. An emerging potential useful method in addition to antigen or DNA detection is the use of immunomagnetic beads to isolate the organism.

Conclusions

Although there are many well validated methods that can be used to detect H pylori, there are nevertheless continued improvements to current tests and the development of new tests, driven by considerations of cost and a search for speed and technical simplicity. An increasing number of patients may be treated by their general practitioners and consequently there is a requirement for a rapid, simple, accurate, near patient test. Furthermore, the emergence of antibiotic resistant isolates has given an impetus to the development of rapid molecular techniques to identify such resistant organisms. The current main areas of development in diagnostic tests are in the use of non-invasive salivary antibody and faecal antigen tests, the use of rapid near patient antibody tests, and the development of DNA immunoassays for the detection of both antibiotic sensitive and resistant H pylori.

References

View Abstract

Footnotes

  • Abbreviations used in this paper:
    ELISA
    enzyme linked immunosorbent assay
    RIBA
    recombinant immunoblot assays
    PUD
    peptic ulcer disease
    UBT
    urea breath test
    RUT
    rapid urease test
    PCR
    polymerase chain reaction

Request permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.