Best Practice & Research Clinical Gastroenterology

Vol. 21, No. 2, pp. 299e313, 2007

doi:10.1016/j.bpg.2006.11.002
available online at http://www.sciencedirect.com

Diagnosis of Helicobacter pylori: Invasive

and non-invasive tests

Chiara Ricci
Lecturer of Internal Medicine
Gastroenterology Unit, University of Brescia, Italy

John Holton
Senior Lecturer in Microbiology
Centre for Infectious Diseases and International Health, Royal Free and University
College London Medical School, UK

Dino Vaira *
Associate Professor of Internal Medicine
Department of Internal Medicine and Gastroenterology, Nuove Patologie, University of Bologna,
Via Massarenti 9, 40138 Bologna, Italy

Helicobacter pylori infection can be diagnosed by invasive techniques requiring endoscopy and biopsy
(e.g. histological examination, culture and rapid urease test) and by non-invasive techniques, such as
serology, the urea breath test, urine/blood or detection of H. pylori antigen in stool specimen. Some
non-invasive tests, such as the urea breath test and the stool antigen test, detect active infection:
these are called ‘active tests’. Non-invasive tests (e.g. serology, urine, near-patient tests) are markers
of exposure to H. pylori but do not indicate if active infection is ongoing; these are ‘passive tests’.
Non-invasive test-and-treat strategies are widely recommended in the primary care setting. The
choice of appropriate test depends on the pre-test probability of infection, the characteristics of
the test being used and its cost-effectiveness.

Key words: antigen test; Helicobacter pylori; invasive tests; non-invasive tests; stool; urea breath
test.

* Corresponding author. Tel.: þ39 051 636 4140; Fax: þ39 051 398 794.
E-mail address: vairadin@med.unibo.it (D. Vaira).
1521-6918/$ - see front matter ª 2006 Elsevier Ltd. All rights reserved.
300 C. Ricci et al

INTRODUCTION

The incidental discovery, in 1983, of a gastric bacterium led to a dramatic change in the
field of gastroenterology.1 Helicobacter pylori infects more than half the global popula-
tion, causing peptic ulcer disease and chronic gastritis; it is also strongly associated
with gastric malignancies. Indeed, it has been classified as a class I carcinogen.2 H. pylori
infection can be diagnosed by invasive techniques (endoscopy with biopsies for histol-
ogy, culture and a rapid urease test) and non-invasive techniques (e.g. serology, the
13
C-urea breath test and the stool antigen test).3
All the available tests seem to be accurate in the diagnosis of H. pylori. However,
a single test (with the exception of culture) is not sufficient to make a diagnosis of
infection. For this reason, the European Guidelines take the gold standard to be
generally represented by at least two different tests.4 Despite this, it is usual in daily
clinical practice to use just one test for diagnosis of infection; choosing the ‘right
test’ is therefore most important.
Any diagnostic test can be characterised by its sensitivity and specificity, and by its
positive and negative predictive values; the latter parameters vary with the prevalence
of the disease in the community. The prevalence of a condition is the proportion of
individuals in a population with the condition. A perfect test will correctly identify
all those who have the disease by recording a positive test (sensitivity 100%). Numer-
ically, this is the number of true positive (TP) divided by the number of TP plus the
false negative (FN): TP/(TP þ FN). A perfect test will also identify all those who do
not have the disease by recording a true negative (TN) test (specificity 100%): TN/
(TN þ FP). For example, if the prevalence of a condition is 40% (i.e. 400/1000 have
the disease) and the test identifies 340/400 persons with the disease correctly, the
test sensitivity is 85%; if the test identifies 480/600 persons without the disease cor-
rectly, the specificity is 80%. If the prevalence of disease was only 10% the sensitivity
and specificity would be unchanged and would detect 85/100 as positive and 720/900
as negative.
There is generally a compromise between sensitivity and specificity in the test char-
acteristics. For diseases that are amenable to treatment, a test with a high sensitivity is
preferable, whereas for those conditions that are untreatable or have implications for
the population at large, a test with a high specificity is required.
However, although they are reported in the literature more often than not, sensi-
tivity and specificity are not particularly helpful; an alternative way to define them is:
the probability of persons with the disease having a positive test (sensitivity) and the
probability of persons without the disease having a negative test (specificity). What is
more relevant is the probability of a person having (or not having) the disease given
a positive (or negative) test. Such a calculation is possible using Bayes’ equation:

Probability of disease with a positive test ¼ ðprevalence  sensitivityÞ

=ðprevalence  sensitivityÞ þ ½ð1  prevalenceÞ  ð1  specificityÞ

The Bayesian approach requires a pretest probability (in this case prevalence) to be
converted into a post-test probability. In Bayes’ theorem, the post-test probability is
proportional to the pretest probability multiplied by the likelihood of the test being
positive. This value is the positive predictive value (PPV) related to the prevalence
of disease in the community. Thus, in populations with a low prevalence of disease,
the PPV will be low even if the sensitivity and specificity are high, and there will be
 Diagnosis of Helicobacter pylori 301

many false-positive results. The converse is true of the negative predictive value
(NPV):

NPV ¼ Specificity  ð1  prevalenceÞ=½ð1  sensitivityÞ  prevalence

þ ½specificity  ð1  prevalenceÞ

The likelihood ratio (the ratio of the probability of a positive result in a person with
the disease compared with the probability of positive results in a person who does not
have the disease) can indicate the value of test.
The choice of test should therefore be based on: (1) the prevalence of the infection
in the population; (2) the symptoms, such as the presence of alarm symptoms; (3) the
likelihood ratio for a positive and negative test; (4) the costs; and (5) the availability of
the tests in the different settings.
Endoscopy was commonly used in diagnostic units but this procedure increases
costs and waiting times. More recently, the ‘test-and-treat’ strategy has been suggested.
This uses a non-invasive test in patients without alarm symptoms (anaemia, weight loss,
etc.), who are aged <55 years and who are not taking non-steroidal anti-inflammatory
drugs (NSAIDs). Endoscopy is still necessary when the patient is >55 years, if there are
alarm symptoms or if there are acute new dyspeptic symptoms.5

INVASIVE TESTS

Endoscopy

Upper gastrointestinal endoscopy is an expensive and unpleasant procedure that car-

ries the risk of haemorrhage and perforation; it has a reported mortality of 0.008% and
a morbidity of 0.432%.6 The European Helicobacter Study Group (EHSG) Maastricht-2
Report recommended that, in the primary care environment, patients who were
<45e50 years of age, were presenting with dyspepsia in the absence of alarm symp-
toms or symptoms of gastro-oesophageal reflux disease (heartburn, regurgitation) and
who were not taking NSAIDs, should be offered diagnosis by a test-and-treat strategy
using the non-invasive urea breath test or the faecal antigen test. A search-and-treat
strategy was recommended for patients with peptic ulcer who were on long-term
acid-suppressive treatment. Endoscopy should be performed in patients <45e50
years who had alarm symptoms and in all patients over this age limit, irrespective of
alarm symptoms. This strategy was endorsed by the EHSG Maastricht-3 guidelines
and supported by a number of publications, including a Cochrane systematic review.7e11
The most significant findings were a reduction in the number of endoscopies and a price
difference of £160 for the test-and-treat compared with £400 for the endoscopy strat-
egy. Endoscopies had a marginal advantage in older patients but even here there was
a definite cost saving. Further recommendations of the EHSG Maastricht-3 guidelines
were that a test-and-treat strategy was optimal in all adult patients with functional
dyspepsia in areas of high Helicobacter prevalence but that its efficacy was less in low
Helicobacter prevalence areas; an option here being empirical acid suppression.
In recent years there has been increasing interest in the goal of endoscopic in-vivo
histology using narrow band imaging, chromoendoscopy and confocal laser endomi-
croscopy, where the mucosal surface and subsurface can be examined in detail for
the presence of characteristic pathological features and the detection of H. pylori.12,13
302 C. Ricci et al

Narrow band imaging endoscopy is based on the principle of splitting white light into
red, green and blue by narrow band filters (thus preventing wavelength overlap); the
wavelengths being reflected from the mucosal surfaces at different depths. The inte-
grated final image has the majority contribution from the blue part of the spectrum
and reveals the surface microvasculature.
Confocal laser endomicroscopy was developed by including a confocal laser micro-
scope in the tip of a video endoscope. The diameter of the insertion tube is 12.8 mm,
indicating a high degree of miniaturisation. Images can be viewed at a resolution of
1024 pixels with a field of view of 500 mm2 and at a depth of 250 mm. A contrast agent
(either acriflavin or fluorescein) has to be used to see the cellular details. Fluorescein
does not stain nuclei (given intravenously) whereas acriflavin (a topical agent) does and
is thus of use in determining nuclear abnormalities and in combination with fluorescein
has also been used to detect the presence of H. pylori on the mucosal surface.14 The
endoscopic appearance of the normal stomach is a regular arrangement of the epithe-
lial cells, gastric pits and the subepithelial capillaries, which give a honeycomb
appearance. In H. pylori gastritis, the collecting venules cannot be seen, the gastric
pits are enlarged with surrounding erythema and the normal capillary network is
lost (Figures 1 and 2). In atrophic gastritis, both the gastric pits and capillary network
are lost. In gastric cancer, there is loss of the capillary network, irregular branched
tubules and loss of polarity and pleomorphism of the nuclei.
Several conditions are characterised by prominent mucosal folds, which can be seen
by endoscopy. In lymphocytic gastritis, which is thought to be an unusual reaction to
H. pylori and is also found in coeliac disease, the surface mucosa can appear normal,
although a typical appearance is varioliform, having prominent rugal folds carrying
nodular elevations with white erosions at the apex and surrounded by a hyperaemic
margin.15 In Menetrier’s disease there are also prominent folds, which, on endosono-
graphy, reveal hyperplasia of the deep mucosa (layer 2).16 The regularity of the epithe-
lial cells, arrangement of the gastric pits and nuclear morphology can thus all be used
to make a real-time pathological diagnosis. Additionally, this advance in endoscopic
technique opens the way for the development of real-time in-vivo sensitivity testing.
Endoscopists should be specially trained in gastric histopathology. A shortcoming of
these endomicroscopic techniques is an inability to view very far below the mucosal
surface. Endoscopy also provides biopsy specimens that can be used in different diag-
nostic tests such as culture, histology, rapid urease test and the polymerase chain
reaction (PCR).

Figure 1. Confocal laser endoscopy: H. pylori gastritis (from Ref. 14).

 Diagnosis of Helicobacter pylori 303

Figure 2. H. pylori gastritis (from Ref. 14).

Histology

Histology can reveal the presence of bacteria as well as the type of inflammation. Many
stains have been used to detect H. pylori, for example, WarthineStarry, Hp silver stain,
Dieterle, Giemsa, Giminez, acridine orange, McMullen and immunostaining. Currently,
the guidelines suggest that at least two stains are used: haematoxylin & eosin to eval-
uate the inflammatory cells and the Giemsa or Genta stain to detect H. pylori. The
Genta stain has the advantage of visualising both the inflammatory cells and H. pylori
by combining a silver stain, haematoxylin & eosin and Alcian blue, although this is tech-
nically complex and uses uranyl nitrate in its original formulation. Overall, the Giemsa
stain is the preferred stain for detecting H. pylori because of its technical simplicity, high
sensitivity and low cost.17,18 Despite the high sensitivity of histology, the site, the num-
bers and the size of the biopsies affect the diagnostic accuracy. Patchy colonisation can
also cause misdiagnosis. Everybody should be aware that although a single biopsy taken
in the lesser curve, close to the angulus, can detect H. pylori presence in more than
90%, the accuracy could be increased with multiple biopsies from the greater curve
and corpus. Generally, specificity is high because of the peculiar morphology and its
close relation to gastric mucosa.19
The histological appearance of gastritis uses the updated Sydney system. This has
a visual analogue scale with semi-quantitative scoring of mild, moderate and marked,
and can score the density of H. pylori, granulocyte (acute gastritis) and mononuclear
cell (chronic gastritis) infiltration, atrophy and intestinal metaplasia.20 There is a marked
infiltration of the lamina propria by granulocytes in acute gastritis; in chronic gastritis
this becomes a predominantly mononuclear infiltration.
In a histological section, H. pylori is recognised by its appearance as a short, curved or
spiral bacillus resting on the epithelial surface or in the mucus layer; it is also found deep in
the gastric pits. Other Helicobacter species, such as H. heilmanii and H. bizzozeroni, are also
detected in the human stomach. H. heilmanii is prevalent in about 0.1% of gastric biopsies.
Its appearance differs from that of H. pylori: it is straight, rather than curved, and is much
longer (about 10 mm), with tight spirals giving it a corkscrew shape. Gastritis associated
with H. heilmanii has a characteristic histology of lymphocyte infiltration into gastric fo-
veolae with lack of mucus depletion. H. heilmanii is a zoonotic infection in humans, being
derived from cats or dogs, and can give rise to chronic gastritis and possible MALT lym-
phoma.21 H. bizzozeroni is similar to in appearance to H. heilmanii.
304 C. Ricci et al

The sensitivity and specificity of histology for the diagnosis of H. pylori varies from
53% to 90%, depending partly on the clinical setting, partly on the density of coloni-
sation and partly on the experience of the histopathologist.
In general, a histological diagnosis can be made in about 90% of cases.22,23 The
average time for a histological diagnosis is 2e3 days, However, this increases when
multiple biopsies are taken, which also increases the processing costs of the biopsies
and the overall costs of the diagnosis. Prior treatment to reduce the numbers of
H. pylori will adversely affect the sensitivity of histology.
In addition to the typical appearance of acute or chronic gastritis associated with
H. pylori or non-pylori Helicobacter, other forms of gastritis that are associated with
Helicobacter infection occur, such as lymphocytic gastritis and Menetrier’s gastritis.
Lymphocytic gastritis is characterised by the accumulation of small lymphocytes in
the surface and foveolae epithelium. A value of >25 intraepithelial lymphocytes/100
epithelial cells is the diagnostic criterion. This is present in about 1% of unselected
endoscopies and in about 4% of patients with chronic gastritis. The histological appear-
ance correlates well with the endomicroscopic appearance of varioliform gastritis.15
In Menetrier’s gastritis, which is thought to be associated with H. pylori, there is a
hyperplastic epithelium with elongated, dilated and tortuous gastric glands penetrating
the muscularis mucosae, cyst formation and an oedematous lamina propria with an in-
flammatory cell infiltrate when in the presence of H. pylori.24
MALT lymphomas, the majority of which are B cell in origin, have been defined as
infiltrates of centrocytoid cells, which have the appearance of plasma cells and which
destroy the foveolae or gastric glands. The lymphomatous cells infiltrate all layers
of the stomach. In low-grade lymphomas, the architecture is maintained and the cell
lineage spans the histological range from a small lymphocyte, to a centrocyte-like
cell, to a monocyte-like B cell, to a lymphoplasma cell to a centroblast cell.25

Culture

Helicobacter can be cultured from gastric biopsies. The colonies are identified by
a Gram stain and biochemical tests. The colonies are Gram negative, urease positive,
oxidase positive and catalase positive. Helicobacter is very fragile out of the gastric en-
vironment: this is why the culture has to be processed as soon as possible. The biop-
sies can be kept in a transport medium (Stuart’s transport medium) for 24 h at 4  C.
Helicobacter are isolated on agar (Columbia or braineheart infusion), generally with
added antibiotics and albumin. The agar plates are incubated in a micro-aerobic envi-
ronment, which is obtained by using a jar with a gas-generating kit for a micro-aerobic
atmosphere (5% oxygen and 5e10% CO2). The plates are incubated for at least 5 days
at 37  C, even though Helicobacter colonies sometimes appear after just 3 days.
Although culture has a high specificity (100%), the sensitivity is often lower. This might
be because: an insufficient number of biopsies was taken, there was a delay in transport-
ing the culture to the laboratory, the culture was exposed to an aerobic environment or
the cultures might not have been recognised as a result of microbiological inexperience.
Moreover, antibiotics, proton pump inhibitors and H2 antagonists must not taken in the
preceding 2 weeks because they also reduce the sensitivity of culture.
Culture tends to be done only in research centres particularly dedicated to H. pylori
infection. However, as the prevalence of antibiotic resistance increases globally there is
a strong argument for performing culture and sensitivity testing after the first treat-
ment failure (to prevent the emergence of double resistance to clarithromycin and
 Diagnosis of Helicobacter pylori 305

metronidazole), and certainly after the second. Indeed, some would argue that it
should be performed at the initial diagnosis in areas of high resistance prevalence.26e28
Also, it has to be emphasised that susceptibility for a full range of the antibiotics can
only be done on culture. When testing for antibiotic sensitivity on a routine basis, the
two most used tests are disc diffusion and the E-test, with the agar dilution test as the
reference. For clarithromycin and amoxycillin there is a good correlation between the
E-test and agar dilution, however, for metronidazole the E-test tends to give higher
MICs compared to the reference method or the disc test.29,30 An additional important
reason for culturing H. pylori is for research purposes, such as correlating specific
virulence characteristics with different clinical outcome and investigation of the molec-
ular events following binding of the organism to gastric epithelial cells.

Rapid urease test

This gastric biopsy test is based on the activity of the H. pylori urease enzyme, which
splits the urea test reagent to form ammonia. Ammonia increases pH, which is de-
tected by the indicator phenol red. Although some of the commensal flora of the oro-
pharynx produce urease, which is swallowed in the saliva, this weaker enzyme is
denaturated rapidly in the acid lumen of the stomach (pH < 2.0). Many commercial
urease tests are available, including gel-based tests (CLOtest, HpFast) paper-based
tests (PyloriTek, ProntoDry HpOne) and liquid-based tests (CPtest, EndoscHp). The
tests give a result in 1 h to 24 h, depending in part on the format of the test and
the number of Helicobacter in the biopsy specimen. In some tests (PyloriTek), the sug-
gested advantage is the built-in positive and negative control with each test strip, and
the fact results are obtained within an hour, as opposed to, for example, the CLOtest,
which takes 24 h. Unbuffered tests can give results in less than 1 h, e.g. the CPtest, and
these liquid tests can be prepared cheaply in-house by adding phenol red to a solution
of urea and adjusting the pH to 6.6. At this pH the indicator is yellow and, on addition
of a positive biopsy, will convert the solution to a red colour. One disadvantage of in-
house tests, apart from quality control, is a short shelf life. All the commercial rapid
urease tests have specificities above 95e100% but the sensitivity is slightly less, at
85e95%.31e36 In comparison to histology, culture and PCR, urease tests are more
rapid, much cheaper and have comparable sensitivity and specificity (except for cul-
ture, which has 100% specificity).
The sensitivity is affected mainly by the number of bacteria present in the biopsy. It
has been calculated that 104 organisms are required for a positive result and a propor-
tion of patients can harbour densities lower than this. Low sensitivity and specificity
are also reported in post-treatment and in bleeding patients, and for these reasons
their use is not advised in these clinical settings.
False-negative urease test can also be obtained in patients with achlorhydria as well
as in patients on proton pump inhibitors because the increased luminal pH can lead to
extremely high pH adjacent to the organism, such that H. pylori is destroyed by the
action of its own urease.37

Molecular tests

Two molecular tests that are available are in-situ hybridisation and the PCR. In-situ
hybridisation has been used to detect the presence of H. pylori in archival specimens
306 C. Ricci et al

and to detect the presence of specific virulence markers; however, it has more often
been used to detect clarithromycin resistance.38e41 The sensitivity and specificity for
the diagnosis of H. pylori infection using in-situ hybridisation with biotinylated probes
has been reported at 95% and 100%.
PCR has been used extensively for the diagnosis of H. pylori from gastric biopsy
specimens, saliva, faeces and archival specimens, as well as for detecting clarithromycin
resistance.42e44 PCR yields information on the presence of potential virulence
markers in the strain, which might have implications for the development of severe
disease or efficacy of eradication. Different primers have been utilised and some
have been developed into commercial kits. Different loci have been used as the target
for the amplification: 16S rRNA; A-, B- and C-urease; flaA; cagA; vacA and heat-shock
protein (hsp). Real-time results can be obtained using light-cycle technology.
PCR detection from faeces is less generally used, partly because cheaper, technically
easier and more reliable tests are available and partly because some faeces are replete
in inhibitors of the PCR reaction. PCR is also been used to test the chemosusceptibil-
ity of Helicobacter to clarithromycin, which is used in the standard eradication regimen,
because resistance to clarithromycin is caused by point mutations (A-G translocation)
in the 23S rRNA gene.45 It is more difficult to detect multiple mutations, although it is
possible with a line probe assay (LiPA) using biotinylated probes for the different mu-
tations or after amplification with a preferential homoduplex formation assay
(PHFA).46,47
Although currently a research tool and not generally used in routine diagnosis,
microarrays might be used diagnostically for H. pylori in the future. The disadvantages
of PCR as a routine test are that it is a technically demanding and expensive test com-
pared to culture, histology and the rapid urease tests. It requires special laboratory
conditions with separate facilities for each stage of the technique and, as it is highly
sensitive, it is subject to false-positive results by contamination. A positive result de-
tected by any of the molecular techniques does not indicate current infection but will
also detect the DNA or dead organisms.48

NON-INVASIVE TESTS

These tests can either be either active or passive. Active tests detect the presence of
H. pylori, i.e. provide evidence of a current infection. The stool test detects the pres-
ence of bacterial antigens of H. pylori in stool; these antigens disappear when H. pylori
infection is cured. The urea breath test is another direct test based on the detection of
urease activity, which is demonstrated in individuals with active H. pylori infection. Pas-
sive tests provide evidence of exposure to H. pylori at some time and do not indicate
whether the infection is currently active. These tests are based on the detection of
antibodies to H. pylori.

Passive tests

Serological tests
There are three main formats for these tests: the most common is the enzyme-linked
immunosorbent assay (ELISA) test, which detects the totality of the immunoglobulin in
a patients’ serum and can detect any of the immunoglobulin isotypes. If the antibody
response to specific antigens is important, the separate ELISA test has to be done
 Diagnosis of Helicobacter pylori 307

for each antigen. This test is usually done in the laboratory because it requires an
ELISA reader.
Latex agglutination tests require only minimal equipment and can detect all the
immunoglobulin isotypes. They are performed more rapidly than an ELISA test and
are commonly used for near-patient tests.
The third type of test is based on Western blotting: the specific antigens are sepa-
rated by gel electrophoresis, transferred to a filter-paper strip and reacted with the
patient’s serum sample. These immunoblots detect all the immunoglobulin isotypes
but have the advantage of detecting specific antigens in one assay. Commercial immu-
noblots are used in near-patient testing as they require minimal apparatus and are rapid.
Since the appearance of the first serological tests for the diagnosis of H. pylori in-
fection, they have become the most frequently used tests in routine practice because
of their accuracy, low cost and availability. Serological tests are based on the detection
of specific anti-H. pylori IgG antibodies in a patient’s serum. Some tests also detect the
presence of IgA in the saliva or IgG in the urine. The advantages of serological test are
that they require no specialised equipment or technique and can be performed by
most hospital or clinic laboratories.
The sensitivity and specificity of the tests depends on the antigen used, the clinical
context, the gold standard used as a comparator and the prevalence of H. pylori in
the community. Overall, the sensitivity has been reported (mostly for the ELISA-based
format) as ranging between 90% and 97% and the specificity as between 50%
and 96%.49e53
The most important disadvantage of serological tests is that they do not distinguish
between active infection and a previous exposure to H. pylori. Antibody levels can per-
sist in the blood of individuals cured of H. pylori infection for long periods of time.
Therefore, as the numbers of patients successfully treated for H. pylori increase in
a population, the prevalence of false-positive tests with serology increases. Recently,
McNulty54 correctly stated that a positive serological test can mean: (1) that the
patient is infected at the time of the test; (2) that the patient was once infected
but, by the time of the test, infection has resolved; or (3) that the test is detecting
non-specific cross-reacting antibodies. Therefore, using a serology-based test, 255
of 2000 patients tested are likely to receive an incorrect diagnosis of active H. pylori
infection and receive inappropriate treatment.54

CagA and vacA antibodies

CagA is a 120-kDa protein of H. pylori with high immunogenicity. Its gene (cagA) is con-
tained in the pathogenicity island (PAI) of the chromosome of H. pylori. Individuals in-
fected with cagA-positive strains of H. pylori tend to have more severe gastritis,
a higher likelihood of developing gastric atrophy and intestinal metaplasia, and a higher
incidence of duodenal ulcer and intestinal-type gastric cancer.12 The vacuolating toxin
(vacA) is an 87-kDa protein and its gene is not located in the pathogenicity island.
However, it is secreted by Helicobacter strains that contain the PAI. The vacuolating
toxin (vacA) is activated at low pH, is resistant to acid and pepsin, and causes vacuo-
lation of gastric cells in cell culture in vitro; in animal models it has been shown to dam-
age mouse gastric epithelium.
Antibody tests still play an important role in studies of pathogenesis and virulence.
Although these tests are of limited value in the clinical setting, they have increased our
understanding of the pathogenesis of disease caused by H. pylori. Antibodies against the
important proteins of H. pylori, cagA and vacA, can be detected using different
308 C. Ricci et al

immunological techniques. After eradication of H. pylori, the antibodies disappear at

different times and some, such as the anti-CagA antibodies, may persist for years. Al-
though the antibody response to cagA can be detected using an ELISA technique, im-
munoblots have a particular advantage in that they are more sensitive and, as
mentioned, can provide evidence of immunological reactions to several antigens all
at once, which might have clinical relevance.49,55

Near-patient tests
Near-patient tests were developed to provide a rapid diagnosis of H. pylori infection in
clinics or physician offices; they are also called office-based serology tests. They are
technically simple to perform and the most convenient of them uses a drop of whole
blood obtained by finger-prick. Other near-patient tests use serum, require venepunc-
ture and use a centrifuge to separate the serum. The results of finger-prick tests can
vary significantly when the flow of blood is poor and there is difficulty in obtaining
a drop of blood; squeezing the finger can express tissue fluid into the blood sample,
thereby changing the concentration of antibody sample being studied. Whole-blood
tests are affected by circulating chylomicrons, which change the permeability of the
blood sample to the various antigens in the membranes used in diagnostic tests. Stud-
ies have reported a lower sensitivity and specificity than was originally suggested; the
mean sensitivity was of 71% and specificity of 87%. A Canadian study of serology found
that 33% of positive near-patient tests in dyspeptic patients in primary care setting
were false positives.56 The current near-patient tests are not recommended for diag-
nosis as they are likely to be false-positive results if the prevalence of the disease is
low, even though the sensitivity may be high.

Tests on saliva and urine

Tests on saliva and urine are attractive because samples can be easily obtained. They
share the same probability as other antibody tests in the serum and have an additional
problem in that the concentration of the antibody is lower than in serum, making de-
tection more difficult. Individual centres have reported promising results (sensitivity
81%, specificity 95%). However, these data were not confirmed in a multicentre trial
using the same urine assay kit, when a sensitivity of 89% and a specificity of 69% were
reported.57

Active tests for the detection H. pylori

These tests are useful for the initial detection of H. pylori and for confirming
eradication.

13
C-urea breath test
H. pylori produce urease, an enzyme that splits urea into ammonia and carbon dioxide.
Indeed, the organism is thought to use urease activity to regulate pH in its micro-
environment. The urea breath test is based on the principle that urease activity is pres-
ent in the stomach of individuals infected with H. pylori. Patients ingest urea labelled
with either 13C or 14C. Hydrolysis of urea occurs within the mucus layer and results
in the production of labelled CO2. The CO2 diffuses into the epithelial blood vessels
and, within a few minutes, the isotopic CO2 appears in the subject’s breath. Labelled
 Diagnosis of Helicobacter pylori 309

urea is usually given to the patient with a test meal to delay gastric emptying and in-
crease contact time with the mucosa. After ingestion of the urea, breath samples are
collected for up to 20 min by exhaling into a CO2-trapping agent (hyamine). Urea
breath test have a very high sensitivity and specificity, ranging from 95% to 97%. How-
ever, the urea breath test might not be reliable when assessing patients who have had
gastric surgery or those who have been on a proton pump inhibitor or ranitidine.58
The 13C urea breath test is similar to the 14C urea breath test except that 13C is
a non-radioactive isotope of 12C and its detection requires a mass spectrometer
rather than a scintillation counter. An alternative detection method is infra-red spec-
troscopy, which is technically simpler and also cheaper than a mass spectrometer.59
Because 13C is a naturally occurring stable isotope, there are no nuclear regulatory
concerns and the test can be used in children and pregnant woman. Since the initial
report of this test, it has been extensively modified, including variations in dose-
labelled urea, sampling time, test meal and cut-off value. The standard urea breath
test uses 75 mg of 13C and was compared to a rapid low-dose (50 mg 13C) test for-
mulated as a tablet also containing citric acid. The test was utilised both as a diagnostic
and to evaluate clearance after eradication of H. pylori. It gave the standard test sen-
sitivity and specificity of 100%. Doses of 13C as low as 10 or 15 mg also gave sensitivity
and specificity of 89e96% and 100%, respectively, and the 15-mg dose was accurate
for assessing eradication.60,61

Stool antigen test

An enzymatic immunoassay (EIA), which detects the presence of H. pylori antigen in
stool specimen, has recently become available. This assay has undergone extensive
testing for the initial diagnosis of H. pylori infection and for confirmation of erad-
ication after treatment. The most widely used test in the assay uses polyclonal anti-
H. pylori-capture antibodies absorbed to microwells. This polyclonal antibody test
has been extensively evaluated in the diagnosis of H. pylori infection before therapy.
Most studies have suggested that the test’s accuracy is similar to that of the urea
breath test in the initial diagnosis of H. pylori infection.62,63 A recent study sug-
gested that the urea breath test and stool antigen test are equally accurate in con-
firming eradication. A positive stool test 7 days after completion of treatment is
predictive of failed eradication.64
Very recently, Gisbert et al, have reported in a metanalysis paper, the superiority of
monoclonal stool antigen compared to polyclonal test, both for the initial diagnosis of
H. pylori and for confirmation of its eradication after treatment. (Table 1).65 According
to the European Guidelines, the monoclonal test and the urea breath test are the two
non-invasive tests recommended for monitoring the success or failure of eradication
treatment.5 A new rapid monoclonal one-step lateral flow immunoassay stool test,

Table 1. Monoclonal test overall results.

Pre-treatment Post-treatment
N ¼ 2499 patients N ¼ 957 patients
Sensitivity 94% 93%
Specificity 97% 96%
Source: Gisbert et al, Am J Gastroenterol 2006;101:1921e30.
310 C. Ricci et al

Figure 3. ImmunoCard Stat! HpSA (from Ref. 66).

taking only 5 min, has been recently been released (Figure 3). Stool is placed in a diluent
vial and then a drop of the liquid is placed in the well of the test. Early results are
promising, with sensitivity and a specificity of 92%.66 This test might be useful as
a near-patient test in primary care settings. It also has a particular advantage in nee-
dle-phobics and children, who are generally averse to venepuncture. In principle, the
test can be amended to pick-out specific proteins, such as cagA or vacA, which could
be used to determine the virulence type of the infecting strain.

COST-EFFECTIVENESS AND STRATEGIES BASED ON PRETEST

PROBABILITY OF INFECTION

The choice of an initial test for H. pylori detection depends on the prevalence of H. pylori
infection and the value placed on increased diagnostic accuracy. A recent cost-benefit
analysis evaluated single test and the sequential testing strategy using costs in the US
healthcare system.67 The serology test had the lowest cost per correct diagnosis at
low (30%), intermediate (60%) and high (90%) prevalence ($90e95/correct diagnosis)
but its diagnostic accuracy was low (80e84%). At low and intermediate prevalence, the
stool test was more accurate (93%), with an average cost of $126e127 per correct
diagnosis and an incremental cost of $336e381 per additional correct diagnosis. ELISA
testing was preferable when prevalence rates were very high (90%) and using a confir-
matory urea breath test when the ELISA test was negative increased the diagnostic
accuracy to 96% with modest incremental costs. If the cost of the breath test was
less than $50, or if the cost of the stool test was greater than $82, breath testing
became preferable to stool testing. In patients with a <60% pretest probability of infec-
tion, as is the case in dyspeptic patients in much of Western world, the stool test
provides increased accuracy with modest incremental costs.
 Diagnosis of Helicobacter pylori 311

Practice point

 the preferred method of diagnosing H. pylori—by non-invasive methods—in

pre- and post-treatment settings are the urea breath test and the stool antigen
test.

REFERENCES

1. Warren JR & Marshall BJ. Unidentified curved bacilli on gastric epithelium in active chronic gastritis.
Lancet 1983; I: 1273e1275.
2. International Agency for Research on Cancer, World Health Organization. Infection with Helicobacter
pylori. In: Schistosomes, Liver Flukes and Helicobacter pylori. Lyon: IARC, 1994, pp. 177e202.
*3. Vaira D & Vakil N. Blood, urine, stool, breath, money and Helicobacter pylori. Gut 2001; 48: 287e289.
*4. Malfertheiner P, Megraud F, O’Morain C et al. Current concepts in the management of H. pylori
infection—the Maastricht 2-2002 consensus report. Aliment Pharmacol Ther 2002; 16: 167e180.
*5. Malfertheiner P, Megraud F, O’Morain C et al. Current concepts in the management of Helicobacter
pylori infection- The Maastricht III Consensus Report. GUT 2007; in press doi:10.1136/gut.2006.101634.
(Dec 14).
6. Quine MA, Bell GD, McCloy RE et al. Prospective audit of upper gastrointestinal endoscopy in two
regions of England: safety, staffing and sedation methods. Gut 2000; 36: 462e467.
7. Heaney A, Collins JS, Watson RG et al. A prospective randomised trial of a test and treat policy versus
endoscopy based management in young Helicobacter pylori positive patients with ulcer-like dyspepsia
referred to a hospital clinic. Gut 1999; 45: 186e190.
8. Jones R, Tait C, Sladen G et al. A trial of a test and treat strategy for Helicobacter pylori positive dys-
peptic patients in general practice. Int J Clin Pract 1999; 53: 413e416.
9. Lassen AM, Pedersen FM, Bytzer J et al. Helicobacter pylori test and eradicate or prompt endoscopy for
management of dyspeptic patients. A randomised controlled trial with one year follow up. Lancet 2000;
356: 455e460.
10. McColl KEL, Murray LS, Gillen D et al. Randomised controlled trial of endoscopy with testing for
Helicobacter pylori compared with non-invasive H. pylori testing alone in the management of dyspepsia.
BMJ 2002; 324: 999e1002.
11. Arents NLA, Thijs JC, van Zwet AA et al. Approach to treatment of dyspepsia in primary care: A rand-
omised trial comparing ‘test and treat’ with prompt endoscopy. Arch Intern Med 2003; 163: 1606e1612.
12. Kiesslich R, Goetz M, Vieth M et al. Confocal laser endomicroscopy. Gastrointest Endosc Clin North Am
2005; 15: 15e31.
13. Uedo N, Ishihara R, Iiishi H et al. A new method of diagnosing gastric intestinal metaplasia: narrow band
imaging with magnifying endoscopy. Endoscopy 2006; 38: 819e824.
14. Kiesslich R, Goetz M, Burg J et al. Diagnosing Helicobacter pylori in vivo by confocal laser endoscopy.
Gastroenterology 2005; 128: 2119e2123.
15. Haot J, Jouret A, Willette M et al. Lymphocytic gastritis-prospective study of its relationship with
varioliform gastritis. Gut 1990; 31: 282e285.
16. Sonqur Y, Okai T, Watanabe H et al. Endosonographic evaluation of giant gastric folds. Gastrointest
Endosc 1995; 41: 468e474.
17. Laine L, Lewin DN, Naritoku W et al. Prospective comparison of the H&E, Giemsa and Genta stains for
the diagnosis of Helicobacter pylori. Gastrointest Endosc 1997; 45: 463e467.
18. Rotimi O, Cairns A, Gray S et al. Histological identification of Helicobacter pylori: comparison of staining
methods. J Clin Pathol 2000; 53: 756e759.
19. Rugge M & Genta RM. Staging and grading of chronic gastritis. Hum Pathol 2005; 36(3): 228e233.
20. Dixon MF, Genta RM, Yardley JH et al. Classification and grading. The updated Sydney System. Interna-
tional Workshop on the Histopathology of Gastritis, Houston 1994. Am J Surg Pathol 1996; 20:
1161e1181.
312 C. Ricci et al

21. Morgner A, Lehn N, Andersen LP et al. Helicobacter heilmannii-associated primary gastric low grade
lymphoma: complete remission after curing the infection. Gastroenterology 2000; 118: 821e828.
22. Cutler AF, Havstad S, Ma CK et al. Accuracy of invasive and non invasive tests to diagnose Helicobacter
pylori infection. Gastroenterology 1995; 109: 136e141.
23. el-Zimaity HM, Graham DY, al-Assi MTet al. Interobserver variation in the histopathological assessment
of Helicobacter pylori gastritis. Hum Pathol 1996; 27: 35e41.
24. Komorowski RA & Caya JG. Hyperplastic gastropathy: clinocopathological correlation. Am J Surg Pathol
1991; 15: 577e585.
25. Isaacson PG. Gastric MALT lymphoma: from concept to cure. Ann Oncol 1999; 10: 637e645.
26. Dore MP, Leandra G, Realdi G et al. Effect of pre-treatment antibiotic resistance to metronidazole and
clarithromycin on outcome of Helicobacter pylori therapy: a meta analytical approach. Dig Dis Sci 2000;
45: 68e76.
27. Romano M, Marma R, Cuomo A et al. Pretreatment antimicrobial susceptibility testing is cost saving in
the eradication of Helicobacter pylori. Clin Gastroenterol Hepatol 2003; 1: 273.
28. Kist M. Impact of German National Reference Centres on monitoring of antibiotic resistance. Int J Med
Microbiol 2006; 296(supplement 41): 55e61.
29. Osato MS, Reddy R, Reddy SG et al. Comparison of the E-test and the NCCLS approved agar dilution
method to detect metronidazole and clarithromycin resistant Helicobacter pylori. Int J Antimicrob Agents
2001; 17: 39e44.
30. Mishra KK, Srivastava S, Garg A et al. Antibiotic susceptibility of Helicobacter pylori clinical
isolates: comparative evaluation of disk diffusion and E test methods. Curr Microbiol 2006; 53:
329e334.
31. Hazell SL, Borody TJ, Gal A et al. Campylobacter pyloridis gastritis, I: detection of urease as a marker of
bacterial colonization and gastritis. Am J Gastroenterol 1987; 82: 292e296.
32. Bermejo F, Boixeda D, Gisbert JP et al. Rapid urease test utility for Helicobacter pylori infection diagnosis
in gastric ulcer disease. Hepatogastroenterology 2002; 49: 572e575.
33. Monteiro L, de Mascarel A, Sarrasqueta AM et al. Diagnosis of Helicobacter pylori infection: noninvasive
methods compared to invasive methods and evaluation of two new tests. Am J Gastroenterol 2001; 96:
353e358.
34. Tseng CA, Wang WM & Wu DC. Comparison of the clinical feasibility of three rapid urease tests in the
diagnosis of Helicobacter pylori infection. Dig Dis Sci 2005; 50: 449e452.
35. Di Bonaventura G, Neri M, Angelucci D et al. Detection of Helicobacter pylori by PCR on gastric biopsy
specimens taken for CP test: comparison with histopathological analysis. Int J Immunopathol Pharmacol
2004; 17: 77e82.
36. Onders RP. Detection methods of Helicobacter pylori: accuracy and costs. Am Surg 1997; 63:
665e668.
37. Basset C, Holton J, Gatta L et al. Helicobacter pylori infection: anything new should we know? Aliment
Pharmacol Ther 2004; 20: 1e10.
38. Ruzsovics A, Molnar B & Tulassay Z. Deoxyribonucleic acid-based diagnostic techniques to detect
Helicobacter pylori. Aliment Pharmacol Ther 2004; 19: 1137e1146.
39. Camorlinga-Ponce M, Romo C, Gonzalez-Valencia G et al. Topographical localisation of CagA positive
and cagA negative Helicobacter pylori strains in the gastric mucosa: an in situ hybridization study. J Clin
Pathol 2004; 57: 822e828.
40. Barrett DM, Faigel DO, Metz DC et al. In situ hybridization for Helicobacter pylori in gastric mucosal
biopsy specimens: Quantitative evaluation of test performance in comparison with the CLO test and
thiazine stain. J Clin Lab Anal 1997; 11: 374e379.
41. Can F, Yilmaz Z, Demirbilek M et al. Diagnosis of Helicobacter pylori infection and determination of
clarithromycin resistance by fluorescence in situ hybridization from formalin fixed paraffin embedded
gastric biopsy specimens. Can J Microbiol 2005; 51: 569e573.
42. Ueda H, Ito M, Eguchi H et al. Development of a novel method to detect Helicobacter pylori cagA
genotype from paraffin embedded materials: comparison between patients with duodenal ulcer and
gastric cancer in young Japanese. Digestion 2006; 73: 47e53.
43. Cellini L, Grande R, Di Campli E et al. Analysis of genetic variability, antimicrobial susceptibility and
virulence markers in Helicobacter pylori identified in Central Italy. Scand J Gastroenterol 2006; 41:
280e287.
 Diagnosis of Helicobacter pylori 313

44. Zsikla V, Hailemariam S, Baumann M et al. Increased rate of Helicobacter pylori infection detected by
PCR in biopsies with chronic gastritis. Am J Surg Pathol 2006; 30: 242e248.
45. Fontana C, Favaro M, Minelli S et al. New site of modification of 23 S rRNA associated with clari-
thromycin resistance of Helicobacter pylori clinical isolates. Antimicrob Agents Chemother 2002; 46:
3765e3769.
46. van Doorn LJ, Glupczynski Y, Kusters JG et al. Accurate prediction of macrolide resistance in H. pylori
by a PCR line probe assay for detection of mutations in the 23S rRNA gene: multicenter validation
study. Antimicrob Agents Chemother 2001; 45: 1500e1504.
*47. De Francesco V, Margiotta M, Zullo A et al. Clarithromycin-resistant genotypes and eradication of
Helicobacter pylori. Ann Intern Med 2006; 144: 94e100.
48. Lisby G. Application of nucleic acid amplification to clinical microbiology. Mol Biotechnol 1999; 12:
75e99.
49. Ekesbo R, Toth E, Fork FT et al. Chronic Helicobacter pylori infection in a population in southern Sweden
analysed by histopathology immunoblot and ELISA serology. Eur J Gastroenterol Hepatol 2006; 18:
589e593.
50. Zuniga-Noriega JR, Bosques-Padilla FJ, Perez-Perez GI et al. Diagnostic utility of invasive tests and
serology for the diagnosis of Helicobacter pylori infection in different clinical presentations. Arch Med
Res 2006; 37: 123e128.
51. Kullavanijaya P, Thong-Ngam D, Hanvivatvong O et al. Analysis of eight different methods for the de-
tection of Helicobacter pylori infection in patients with dyspepsia. J Gastroenterol Hepatol 2004; 19:
1392e1396.
52. Garza-Gonzales E, Bosques-Padilla FJ, Tijerina-Menchaca R et al. Comparison of endoscopy based and
serum based methods for the diagnosis of Helicobacter pylori. Can J Gastroenterol 2003; 17: 101e106.
53. Sternberg A, Coscas D, Wagner Y et al. Comparison of various Helicobacter pylori detection methods:
serology, histology and bacteriology. Isr J Med Sci 1997; 33: 160e163.
*54. McNulty C, Teare L, Owen R et al. Test and treat for dyspepsiaebut which test? BMJ 2005; 330:
105e106.
55. Lu CY, Kuo CH, Lo YC et al. The best method of detecting prior Helicobacter pylori infection. World
J Gastroenterol 2005; 11: 5672e5676.
56. Chiba N, van Zanten S, Sinclair P et al. Treating Helicobacter pylori infection in primary care patients with
uninvestigated dyspepsia: the Canadian adult dyspepsia empiric treatment Helicobacter pylori positive
(CADET-Hp) randomised controlled trial. BMJ 2002; 324: 1012.
57. Leodolter A, Vaira D, Bazzoli F et al. European multicentre validation trial of two new non-invasive tests
for the detection of H. pylori antibodies: urine-based ELISA and rapid urine test. Aliment Pharmacol Ther
2003; 18: 927e931.
58. Gatta L, Vakil N, Ricci C et al. Effect of proton pump inhibitors and antacid therapy on 13C-urea breath
tests and stool test for Helicobacter pylori infection. Am J Gastroenterol 2004; 99: 823e829.
59. Taniquchi Y, Kimura K, Sohara H et al. Simple 13-C urea breath test with infra red spectrophotometer.
J Gastroenterol 1996; 31(supplement 9): 37e40.
60. Gatta L, Vakil N, Ricci C et al. A rapid, low dose 13C Urea tablet for the detection of Helicobacter pylori
infection before and after treatment. Aliment Pharmacol Ther 2003; 17: 793e798.
*61. Gatta L, Ricci C, Tampieri A et al. Accuracy of breath tests using low doses of 13C-urea to diagnose
Helicobacter pylori infection: a randomised controlled trial. Gut 2006; 55: 457e462.
*62. Vaira D, Malfertheiner P, Megraud F et al. Diagnosis of Helicobacter pylori infection using a novel, non-
invasive antigen based assay in a European multicentre study. Lancet 1999; 354: 30e33.
*63. Vaira D, Malfertheiner P, Mégraud F et al. Non invasive antigen based assay for assessing Helicobacter
pylori eradication. A European multicentre study. Am J Gastroenterol 2000; 95: 925e929.
*64. Vaira D, Vakil N, Menegatti M et al. The stool antigen H test for the detection of Helicobacter pylori after
eradication therapy. Ann Intern M 2002; 136: 280e287.
*65. Gisbert JP, de la Morena F & Abraira V. Accuracy of monoclonal stool antigen test for the diagnosis of
H. pylori infection: a systematic review and meta-analysis. Am J Gastroenterol 2006; 101: 1921e1930.
66. Gatta L, Perna F, Ricci C et al. A rapid immunochromatographic assay for Helicobacter pylori in stool
before and after treatment. Aliment Pharmacol Ther 2004; 20: 469e474.
67. Vakil N, Rhew D, Sool A et al. The cost-effectiveness of diagnostic testing strategies for Helicobacter
pylori. Am J Gastroenterol 2000; 95(7): 1691e1698.