Iron Deficiency Anemia
A number of studies have suggested a potential association
between unexplained iron deficiency anemia and H. pylori
infection. The explanation most commonly offered for this
relationship is based upon the development of H. pylori-
associated chronic pangastritis with resultant achlorhydria
and reduced ascorbic acid secretion leading to reduced in-
testinal iron absorption. Other potential explanations for an
association between iron deficiency and H. pylori include oc-
cult blood loss from erosive gastritis and sequestration and
utilization of iron by the organism (47).
Recent large studies from North America have reported
H. pylori infection was an independent risk factor for iron
deficiency anemia in 688 school-aged children from Alaska
(48) and 7,462 children, adolescents, and adults from the
United States (49). In the study by Cardenas and colleagues,
H. pylori infection was associated with an increased risk of
iron deficiency anemia (OR 2.6, 95% CI 1.5–4.6). There is
emerging evidence to suggest that eradication of H. pylori can
improve iron deficiency anemia (50–52) though this remains
controversial. A recent unblinded study in 219 H. pylori-
infected children (7–11 yr) with pretreatment iron deficiency
from Alaska found no difference in the likelihood of iron
deficiency or anemia at 2 months or 14 months following a
6-wk course of oral iron and antibiotics or no antibiotics (53).
The available data support an association between H.
pylori infection and iron deficiency but do not prove cause
and effect. Further properly designed, adequately powered
randomized trials are needed to assess whether H. pylori
eradication offers benefit to patients with unexplained iron
deficiency anemia.
Prevention of Gastric Cancer
Whether curing H. pylori infection can reduce the risk of
developing gastric adenocarcinoma remains unknown (54).
However, there have been a number of recent studies that have
evaluated the effect of H. pylori eradication on surrogate out-
comes such as the severity and distribution of gastritis and
gastric preneoplastic lesions (multifocal atrophic gastritis,
intestinal metaplasia, or dysplasia) (55–58). In a random-
ized, placebo-controlled trial, Leung et al. followed 435 H.
pylori-infected patients for 5 yr after a course of anti-H. pylori
therapy or placebo. In a multiple logistic regression analysis,
1812
Chey and Wong
they observed that persistent H. pylori infection (OR 2.13,
95% CI 1.41–3.24), age
>45 yr (OR 1.92, 95% CI 1.18–
3.11), alcohol consumption (OR 1.67, 95% CI 1.07–2.62),
and drinking local well water (OR 1.74, 95% CI 1.13–2.67)
were independent risk factors associated with intestinal meta-
plasia progression. They concluded that H. pylori eradication
was protective against progression of premalignant gastric le-
sions in their Chinese population study (55). In a study from
Columbia, 795 adults with preneoplastic gastric lesions were
randomized to anti-H. pylori therapy or antioxidants and were
followed with serial endoscopies over 12 yr. Multivariate
analysis revealed a significant regression in histopathology
score as a function of the square of time without H. pylori
infection. Further, patients treated for H. pylori were 13.7%
less likely to experience progression of preneoplastic gas-
tric lesions (57). Wong et al. recruited 1,630 asymptomatic
H. pylori-infected subjects in a high-risk region of China,
and randomly allocated them to H. pylori therapy or placebo,
after which they were followed for 7.5 yr. They reported that
gastric cancer developed in 18 cases. There was an abso-
lute reduction in gastric cancer incidence in subjects who
received H. pylori eradication therapy when compared with
placebo, which was not statistically significant (37% reduc-
tion, P
= 0.33). However, in a subgroup of H. pylori carriers
without precancerous lesions at index endoscopy, the inci-
dence of gastric cancer was significantly lower in subjects
receiving eradication therapy than in those receiving placebo
(P
= 0.02). This study supports the possibility that H. pylori
eradication may reduce the risk of developing gastric cancer
in individuals without precancerous lesions from high-risk
Table 2.
Diagnostic Testing for Helicobacter pylori
Endoscopic Testing
Advantages
Disadvantages
∗
1. Histology
Excellent sensitivity and specificity
Expensive and requires infrastructure and
trained personnel
∗
2. Rapid urease testing
Inexpensive and provides rapid results. Excellent
specificity and very good sensitivity in properly
selected patients
Sensitivity significantly reduced in the
posttreatment setting
∗
3. Culture
Excellent specificity. Allows determination of
antibiotic sensitivities
Expensive, difficult to perform, and not widely
available. Only marginal sensitivity
∗
4. Polymerase chain reaction
Excellent sensitivity and specificity. Allows
determination of antibiotic sensitivities
Methodology not standardized across
laboratories and not widely available
Nonendoscopic Testing
Advantages
Disadvantages
1. Antibody testing (quantitative
and qualitative)
Inexpensive, widely available, very good NPV
PPV dependent upon background H. pylori
prevalence. Not recommended after H. pylori
therapy
∗
2. Urea breath tests (
13
C and
14
C)
Identifies active H. pylori infection. Excellent PPV
and NPV regardless of H. pylori prevalence.
Useful before and after
H. pylori therapy
Reimbursement and availability remain
inconsistent
∗
3. Fecal antigen test
Identifies active H. pylori infection. Excellent
positive and negative predictive values regardless
of H. pylori prevalence. Useful before and after H.
pylori therapy
Polyclonal test less well validated than the UBT
in the posttreatment setting. Monoclonal test
appears reliable before and after antibiotic
therapy. Unpleasantness associated with
collecting stool
∗
The sensitivity of all endoscopic and nonendoscopic tests that identify active H. pylori infection is reduced by the recent use of PPIs, bismuth, or antibiotics
PPI
= proton pump inhibitor; PPV = positive predictive value; NPV = negative predictive value; UBT = urea breath test.
populations (58). No such evidence is available from regions
of the world where gastric cancer is rare, such as the United
States.
A recent international working group reviewed the litera-
ture addressing this topic. The majority of the scientific task
force favored testing and treating H. pylori in first-degree rel-
atives of gastric cancer patients. The task force also endorsed
the evaluation of the chemopreventive benefits for gastric
malignancy with a more general screen and treat strategy
in populations with a high incidence of H. pylori-associated
diseases (54).
DIAGNOSIS OF H. PYLORI INFECTION
• Testing for H. pylori should only be performed if the clin-
ician plans to offer treatment for positive results.
• Deciding which test to use in which situation relies heav-
ily upon whether a patient requires evaluation with upper
endoscopy and an understanding of the strengths, weak-
nesses, and costs of the individual tests.
The methods of diagnostic testing for H. pylori can be
divided into those that do and those that do not require en-
doscopy. Table 2 provides a list of the available diagnostic
tests for H. pylori. There is no single test that can be consid-
ered the gold standard for the diagnosis of H. pylori. Rather,
the most appropriate test for any specific situation will be
influenced by the clinical circumstances, the pretest proba-
bility of infection, as well as the availability and costs of the
individual diagnostic tests.
Guideline on the Management of Helicobacter pylori Infection
1813
Endoscopic Diagnostic Tests
• In patients who have not been on a PPI within 1–2 wk
or an antibiotic or bismuth within 4 wk of endoscopy, the
rapid urease test (RUT) provides an accurate, inexpensive
means of identifying H. pylori.
• For patients who have been taking a PPI, antibiotics, or
bismuth, endoscopic testing for H. pylori should include
biopsies from the gastric body and antrum for histology
with or without rapid urease testing.
• Though culture or polymerase chain reaction (PCR) are
the primary means by which antibiotic sensitivities can
be determined, neither is widely available for clinical use
in the United States and therefore, cannot be routinely
recommended.
There are presently four biopsy-based diagnostic methods
for H. pylori infection. These include the RUT, histology,
culture, and PCR.
Rapid Urease Testing
The RUT identifies active H. pylori infection through the or-
ganism’s urease activity. Gastric biopsies are obtained and
placed into an agar gel or on a reaction strip containing urea,
a buffer, and a pH-sensitive indicator. In the presence of
H. pylori’s urease, urea is metabolized to ammonia and bicar-
bonate leading to a pH increase in the microenvironment of
the organism. A change in color of the pH sensitive indica-
tor signifies the presence of active infection. Commercially
available kits yield results in 1–24 h.
There are a number of commercially available RUT kits in
the United States including the CLOtest, HpFast, HUT-test,
Pronto Dry, and Pyloritek with overall pretreatment sensitiv-
ities of
>90% and specificities of >95% (59, 60). Though
the overall performance of the different tests is comparable,
there are some practical differences between the individual
tests (61).
Medications that reduce the density and/or urease activity
of H. pylori, such as bismuth-containing compounds, antibi-
otics, or PPIs, can decrease the sensitivity of the RUT by up to
25% (59). Though controversial, acute ulcer bleeding at the
time of testing may decrease the sensitivity and negative pre-
dictive value of the RUT (62–66). As a result of the patchy
distribution of H. pylori infection after antibiotics or PPIs,
it is recommended that biopsies for the RUT be obtained
from two sites, the body at the gastric anglularis and greater
curvature of the antrum (67). The simplicity, low cost, and
relatively rapid results make the RUT a practical and cost-
effective means of testing for H. pylori in patients not taking
antibiotics, bismuth, or PPIs who require upper endoscopy.
Unfortunately, the usefulness of the RUT in routine clinical
practice has been compromised by the widespread use of PPIs
as an empiric treatment for upper GI symptoms. As such,
the RUT can rarely be used as a sole means of identifying
H. pylori infection. More commonly, the RUT is combined
with other endoscopic or nonendoscopic modalities to estab-
lish the presence or absence of this infection. No studies have
been performed to define the duration of a PPI’s deleterious
effects on the sensitivity of the RUT. Data with the urea breath
test (UBT) suggest that PPI therapy can cause false-negative
test results for 1–2 wk (68, 69). As the UBT and RUT rely
upon the identification of H. pylori’s urease activity, it is rea-
sonable to suggest that PPIs should be withheld for 1–2 wk
before performance of the RUT. In situations where a patient
has not taken a PPI for a period of 1–2 wk before their proce-
dure, the sensitivity of the RUT is likely sufficient to justify
its use as a single test for H. pylori.
Histology
Histology has been considered by some to be the gold stan-
dard for detection of H. pylori (70). Unfortunately, histology
is an imperfect gold standard as the detection of H. pylori
relies upon a number of issues including the site, number,
and size of gastric biopsies, method of staining, and the level
of experience of the examining pathologist (70). A signifi-
cant advantage of histology over other diagnostic methods is
the ability to evaluate for pathologic changes associated with
H. pylori infection such as inflammation, atrophy, intestinal
metaplasia, and malignancy (71). In fact, some have argued
that type B chronic gastritis (nonatrophic diffuse antral gastri-
tis or atrophic pangastritis) can be used as a surrogate marker
for the infection when organisms are not identified (72). Cer-
tainly the absence of chronic gastritis is a potent negative
predictor for the presence of H. pylori infection.
As the prevalence and density of H. pylori varies through-
out the stomach, particularly in the face of medications that
may reduce the density of H. pylori, multiple biopsies are
needed for accurate diagnosis. It is therefore recommended
that a minimum of three biopsies be obtained, one from the
anglularis, one from the greater curvature of the corpus, and
one from the greater curvature of the antrum, to maximize the
diagnostic yield of histology (70). A recent study found that
the addition of corpus biopsies to antral biopsies increased
the detection of H. pylori infection by
∼10% when com-
pared with antral biopsies alone (73). Similar to the RUT,
the sensitivity of histology is significantly affected by the use
of medications such as bismuth, antibiotics, and PPIs (67).
Although widely available and capable of achieving sensitiv-
ity and specificity of
>95%, the cost and need for properly
trained personnel are limitations of histology in clinical prac-
tice.
Culture
Culture is another highly specific method for identifying ac-
tive H. pylori infection. Conceptually, culture is attractive
because it not only provides a means by which to identify
infection, but also allows characterization of antimicrobial
sensitivities (74). Unfortunately, culture is not as sensitive
as RUT or histology (75, 76). Furthermore, culturing tech-
niques for H. pylori are demanding and costly and as a con-
sequence, only available in a limited number of clinical lab-
oratories. Nonculture-based means of determining antibiotic
1814
Chey and Wong
resistance are being developed but have not been adequately
standardized and are not widely available.
Polymerase Chain Reaction
PCR is a DNA amplification technique that utilizes the rapid
production of multiple copies of a target DNA sequence to
identify H. pylori. This testing method is highly specific and
may be more sensitive than other biopsy-based diagnostic
techniques. A recent study found that PCR was able to de-
tect H. pylori in approximately 20% of gastric biopsies with
chronic gastritis but no identifiable organisms by histology
(77). PCR also provides a means of identifying mutations
associated with antimicrobial resistance (78–80). Although
presently restricted to the research arena, this method may
some day provide a practical, reproducible method for an-
tibiotic sensitivity testing, organism typing, and organism
virulence testing (81).
Nonendoscopic Diagnostic Tests
• Antibody testing is inexpensive and widely available but
poor PPV in populations with a low prevalence of H. pylori
infection limits its usefulness in clinical practice.
• The UBTs and fecal antigen tests provide reliable means
of identifying active H. pylori infection before antibiotic
therapy.
• The UBT is the most reliable nonendoscopic test to docu-
ment eradication of H. pylori infection.
• The monclonal fecal antigen test provides another nonen-
doscopic means of establishing H. pylori cure after antibi-
otic treatment.
• Testing to prove H. pylori eradication appears to be most
accurate if performed at least 4 wk after the completion of
antibiotic therapy.
There are currently three nonendoscopic diagnostic testing
methods for H. pylori infection. Antibody testing identifies an
immunological reaction to the infection while the nonendo-
scopic urease tests and fecal antigen test identify the presence
of active H. pylori infection.
Antibody Tests
Antibody testing relies upon the detection of IgG antibodies
specific to H. pylori in serum, whole blood, or urine. IgG an-
tibodies to H. pylori typically become present approximately
21 days after infection and can remain present long after
eradication (82). Antibodies to H. pylori can be quantita-
tively assessed using enzyme-linked immunosorbent assay
(ELISA) and latex agglutination techniques or qualitatively
assessed using office-based kits. The advantages of the an-
tibody tests are their low cost, widespread availability, and
rapid results. Unfortunately, several factors limit the useful-
ness of antibody testing in clinical practice. A meta-analysis
evaluated the performance characteristics of several commer-
cially available quantitative serological assays and found their
overall sensitivity and specificity to be 85% and 79%, re-
spectively, with no differences between the different assays
Figure 1.
Effect of H. pylori prevalence on the positive predictive
value (PPV) of antibody testing (where sensitivity
= 85% and speci-
ficity
= 79%) (144).
(83). Three of the qualitative whole blood antibody kits were
directly compared in another study demonstrating sensitivi-
ties ranging from 76% to 84% and specificities of 79–90%
(84). In general, performance characteristics for the quali-
tative office-based tests have been more variable than those
yielded by the quantitative tests. It is very important to under-
stand that the PPV of antibody testing is greatly influenced by
the prevalence of H. pylori infection (85) (Fig. 1). This issue
will be further discussed in the section addressing the use of
diagnostic testing in clinical practice. Further, antibody tests
developed using antigens from one region of the world may
not perform well when applied to patients in another part of
the world suggesting that local validation may be necessary
(75, 86). Finally, antibody tests are of little benefit in docu-
menting eradication as results can remain positive for years
following successful cure of the infection (82).
Urea Breath Tests
The UBT, like the RUT, identifies active H. pylori infection
by way of the organism’s urease activity. In the presence of
H. pylori, the ingestion of urea, labeled with either the non-
radioactive isotope
13
C or the radioactive isotope
14
C, results
in production of labeled CO
2
, which can be quantitated in
expired breath (87–90). Although the amount of radiation in
the
14
C UBT is less than daily background radiation expo-
sure (88), the
13
C test is preferred in children and pregnant
females (87). Overall, the performance characteristics of both
tests are similar with sensitivity and specificity typically ex-
ceeding 95% in most studies (87, 88). Test reproducibility has
been found to be excellent (89). The UBT also provides an
accurate means of posttreatment testing (90–93). Most tests
utilize a citrate test meal (50–75 mg), which is administered
before the labeled urea (87). A urease blood test, which relies
upon the detection of labeled bicarbonate in a blood sam-
ple, also reliably identifies active H. pylori infection before
and after treatment (94, 95). As the nonendoscopic urease
tests rely upon the identification of H. pylori’s robust urease
activity, test sensitivity is decreased by medications that re-
duce organism density or urease activity, including bismuth
containing compounds, antibiotics, and PPIs. It is currently
recommended that bismuth and antibiotics be withheld for at
least 28 days and a PPI for 7–14 days prior to the UBT (68,
Guideline on the Management of Helicobacter pylori Infection
1815
Table 3.
Performance Characteristics of the Fecal Antigen Test (95)
# Studies /
# Patients Sensitivity Specificity PPV NPV
Pretreatment
Polyclonol
89/10,858
91
93
92
87
Monoclonol
8/1,399
96
97
96
97
Posttreatment
Polyclonol
39/3,147
86
92
76
93
Monoclonol
6/418
95
97
91
98
PPV
= positive predictive value; NPV = negative predictive value.
69, 96). It is controversial whether H
2
RAs affect the sensi-
tivity of the UBT (97–99) though many laboratories recom-
mend withholding these drugs for 24–48 h before the UBT.
Antacids do not appear to affect the accuracy of the UBT
(100). Aside from the issues just discussed, other factors af-
fecting the acceptance of the UBT in clinical practice include
the need for infrastructure to perform the test, the need for
a patient to attend an additional outpatient visit to undergo
the test, and cost. At current levels of reimbursement in the
United States, the UBT is more costly than the antibody tests
or fecal antigen test. The expense of the UBT is largely driven
by equipment costs and the cost of labeled urea. UBTs using
lower dose
13
C, which have recently been found to yield ex-
cellent performance characteristics, may in part address this
issue (101).
Fecal Antigen Test
The fecal antigen test (FAT) identifies H. pylori antigen in
the stool by enzyme immunoassay with the use of polyclonal
anti-H. pylori antibody. Recently, a stool test utilizing a mon-
oclonal anti-H. pylori antibody has been evaluated (102, 103).
As both tests detect bacterial antigen(s) suggestive of ongo-
ing infection, they can be used to screen for infection and
as a means of establishing cure following therapy. A recent
systematic review (102) reported performance characteris-
tics of the FAT before and after eradication therapy (Table 3).
While this analysis demonstrated excellent sensitivity, speci-
ficity, positive and negative predictive values for the poly-
clonal test before treatment, sensitivity and PPV were less
satisfactory after treatment. On the other hand, the mono-
clonal test yielded sensitivity, specificity, and predictive val-
ues greater than 90% before and after treatment. The precise
explanations for the differences in accuracy between the poly-
clonal and monoclonal tests remain unclear but may have to
do with the need for intraperitoneal injection of H. pylori
antigens into rabbits to produce antibodies for the polyclonal
assay (102). The FAT has been approved by the U.S. Food and
Drug Administration and endorsed by the European “Maas-
tricht 2–2000 Consensus Report” as an alternative means of
establishing H. pylori cure to urea breath testing (104). Recent
studies indicate that the FAT may be effective in confirming
eradication as early as 14 days after treatment (105, 106).
However, there is evidence to suggest that the FAT should be
done more than 4 wk and perhaps as long as 8–12 wk after
treatment of H. pylori (102).
When testing for H. pylori in populations with a low pretest
probability of infection, the FAT provides greater accuracy
than serologic testing with only a modest increase in incre-
mental costs (107). Similar to the UBT, the sensitivity of the
FAT is affected by the recent use of bismuth compounds, an-
tibiotics, and PPIs (108, 109). Recent studies also suggest
that the specificity of the FAT is reduced in the setting of
bleeding PUD and, for this reason, should not be the sole
diagnostic test employed in this setting (110–113). Although
the FAT is simple to administer and perform, issues slowing
its widespread use include the unpleasantness of handling
and storing stool, limited availability, and variable state-to-
state reimbursement. The development of in-office stool tests
is under way and may improve upon some of the practical
limitations of the currently available tests (102). At present,
in-office tests have not been adequately validated in clinical
trials.
Based upon the available data, it is reasonable to conclude
that the FAT can be used interchangeably with the UBT to
identify H. pylori before antibiotic therapy. The polyclonal
FAT has been less well validated than the UBT in the post-
treatment setting. Compared with the polyclonal test, the
monoclonal FAT appears to provide a more reliable means
of proving H. pylori eradication.
H. PYLORI TESTING IN CLINICAL PRACTICE
Testing When There Is a Need for Endoscopy
If endoscopy is necessary based upon the patient’s clinical
presentation, biopsy-based endoscopic tests are most appro-
priate. Provided the patient has not been on recent bismuth,
antibiotics, or a PPI, the RUT offers the desirable combination
of accuracy and low cost. If there are mucosal abnormalities
identified at the time of endoscopy, which require further his-
tologic evaluation, biopsies should be obtained for histology.
Unfortunately, most patients referred for upper endoscopy
are taking acid-suppressive agents such as a PPI or H
2
RA
or have recently received drugs that can suppress H. pylori
(antibiotics or bismuth). In such patients, it is appropriate to
obtain biopsies for histology with or without RUT or plan
testing with a UBT or FAT at a later date after withholding
the offending agents for an appropriate period of time.
In the setting of an active ulcer bleed, there are case se-
ries and cohort studies that suggest that the sensitivity of the
RUT and, to a lesser extent, histology may be reduced (62,
63, 114, 115). These studies suggest that although positive
results reliably identify the presence of H. pylori infection,
the likelihood of false-negative results may be increased in
the setting of acute upper gastrointestinal bleeding. A recent
prospective cohort study from the United States did not con-
firm findings from previous studies (65). In this study, 61
patients with variceal hemorrhage underwent biopsy-based
H. pylori testing during an initial endoscopy for acute bleed-
ing and again 1 month later. The sensitivities of RUT and
histology performed during acute bleeding and 1 month later
were not significantly different. However, it is notable that
1816
Chey and Wong
the sensitivity of the RUT in this study was relatively low
at both time points (initial RUT
= 79%, follow-up RUT =
71%). Regardless of which results one chooses to believe, it
is important to emphasize that a positive RUT indicates the
presence of active H. pylori infection. On the other hand, a
negative RUT and/or histology in the setting of acute upper
GI bleeding should be confirmed with another test. An anti-
body test provides a reasonably sensitive nonendoscopic test-
ing option. In this setting, because the pretest probability of
H. pylori infection is high in a patient with an ulcer, the PPV
of an antibody test is reasonably high (Fig. 1). Alternatively,
a patient can undergo a UBT or FAT at a later date after
withholding medications that can negatively affect the sensi-
tivity of these tests for an appropriate period of time. Recent
work suggests that engaging in such a practice significantly
increases the detection of H. pylori infection in patients with
recent ulcer bleeding. A recent retrospective study from Spain
found that 57 of 72 (79%) patients with ulcer bleeding and no
evidence of H. pylori on emergency endoscopy had a positive
“delayed” UBT (116).
Testing in Patients With Uninvestigated Dyspepsia
Primary care providers are frequently asked to evaluate
and treat patients with uninvestigated dyspepsia. The test-
and-treat strategy for H. pylori has been endorsed for the
management of uninvestigated dyspepsia by a number of
organizations, including the American Gastroenterological
Association (117) and the American College of Gastroen-
terology (15). For a detailed discussion regarding H. pylori
testing in patients with uninvestigated dyspepsia, the reader
is referred to these recent publications (15, 117). Both docu-
ments emphasize that in regions where the prevalence of H.
pylori infection is high, such as urban areas or communities
with large immigrant populations, the PPV of antibody test-
ing is reasonably good and therefore provides an acceptable
means of screening for H. pylori infection. However, in re-
gions where H. pylori prevalence is low, the PPV of antibody
testing is poor (85). From a pragmatic standpoint, this means
that if a physician practices in a community with an H. pylori
prevalence of less than
∼20%, as is the case in much of the
United States, though a negative antibody test suggests the
absence of infection, a positive test is no better than a coin
toss in predicting the presence of active infection (Fig. 1). As
such, in low prevalence populations, antibody tests should
be avoided altogether or positive results should be confirmed
with a test that identifies active infection such as the UBT or
FAT prior to initiating eradication therapy (117, 118).
Testing to Prove Eradication After Antibiotic Therapy
In an ideal world, all patients treated for H. pylori infection
would undergo testing to prove eradication of the infection.
Unfortunately, universal posttreatment testing is neither prac-
tical nor cost-effective. Since publication of the last ACG
guideline on H. pylori infection (3), the accepted indications
for testing to prove eradication after antibiotic therapy, largely
based upon expert consensus, have broadened to include:
• Any patient with an H. pylori-associated ulcer.
• Individuals with persistent dyspeptic symptoms despite
the test-and-treat strategy.
• Those with H. pylori-associated MALT lymphoma.
• Individuals who have undergone resection of early gastric
cancer.
When confirmation of eradication is necessary, testing
should generally be performed no sooner than 4 wk after
the completion of treatment. Because of its high cost, endo-
scopic tests should only be used if endoscopy is clinically
indicated for other reasons. If testing to prove eradication
were performed in the setting of endoscopy, most would ad-
vocate using histology or the combination of histology and
RUT as RUT alone has reduced sensitivity in the posttreat-
ment setting (119). When endoscopic follow-up is unneces-
sary, testing to prove eradication of H. pylori infection is best
accomplished with the UBT. The FAT provides an alterna-
tive means of establishing eradication though, as has already
been discussed, the timing and reliability of this test have
not been as clearly demonstrated as for the UBT. Because
antibody tests can remain positive for prolonged periods fol-
lowing successful cure of H. pylori infection, they should be
avoided in the posttreatment setting. If antibody testing is
performed in the posttreatment setting, only a negative re-
sult is reliable. A positive result should be confirmed with a
UBT or FAT before offering antibiotic therapy for presumed
persistent infection.
TREATMENT OF H. PYLORI INFECTION
Primary Treatment of H. pylori Infection
• In the United States, the recommended primary therapies
for H. pylori infection include: a PPI, clarithromycin, and
amoxicillin, or metronidazole (clarithromycin-based triple
therapy) for 14 days or a PPI or H
2
RA, bismuth, metron-
idazole, and tetracycline (bismuth quadruple therapy) for
10–14 days.
• Sequential therapy consisting of a PPI and amoxicillin for
5 days followed by a PPI, clarithromycin, and tinidazole
for an additional 5 days may provide an alternative to
clarithromycin-based triple or bismuth quadruple therapy
but requires validation within the United States before it
can be recommended as a first-line therapy.
The first course of therapy offers the greatest likelihood of
eradicating H. pylori infection. Subsequent treatment trials,
particularly if the same antibiotics are utilized or if the patient
has been previously exposed to any antibiotics contained in
the treatment regimen, are less likely to achieve a success-
ful outcome. As such, it is important to only use treatment
regimens for which there is evidence of proven effectiveness
(120).
In the United States, the recommended primary therapies
for H. pylori infection include: a PPI, clarithromycin, and
amoxicillin or metronidazole (clarithromycin-based triple
therapy) or a PPI or H
2
RA, bismuth, metronidazole, and tetra-
Guideline on the Management of Helicobacter pylori Infection
1817
Table 4.
First-Line Regimens for Helicobacter pylori Eradication
Regimen
Duration Eradication Rates
Comments
Standard dose PPI b.i.d. (esomeprazole is q.d.),
clarithromycin 500 mg b.i.d., amoxicillin 1,000 mg b.i.d.
10–14
70–85%
Consider in nonpenicillin allergic patients who
have not previously received a macrolide
Standard dose PPI b.i.d., clarithromycin 500 mg b.i.d.
metronidazole 500 mg b.i.d.
10–14
70–85%
Consider in penicillin allergic patients who
have not previously received a macrolide or
are unable to tolerate bismuth quadruple
therapy
Bismuth subsalicylate 525 mg p.o. q.i.d. metronidazole
250 mg p.o. q.i.d., tetracycline 500 mg p.o. q.i.d.,
ranitidine 150 mg p.o. b.i.d. or standard dose
PPI q.d. to b.i.d.
10–14
75–90%
Consider in penicillin allergic patients
PPI
+ amoxicillin 1 g b.i.d. followed by:
5
>90%
Requires validation in North America
PPI, clarithromycin 500 mg, tinidazole 500 mg b.i.d.
5
PPI
= proton pump inhibitor; pcn = penicillin; p.o. = orally; q.d. = daily; b.i.d. = twice daily; t.i.d. = three times daily; q.i.d. = four times daily.
∗
Standard dosages for PPIs are as follows:
lansoprazole 30 mg p.o., omeprazole 20 mg p.o., pantoprazole 40 mg p.o., rabeprazole 20 mg p.o., esomeprazole 40 mg p.o.
Note: the above recommended treatments are not all FDA approved. The FDA approved regimens are as follows:
1. Bismuth 525 mg q.i.d.
+ metronidazole 250 mg q.i.d. + tetracycline 500 mg q.i.d. × 2 wk + H
2
RA as directed
× 4 wk.
2. Lansoprazole 30 mg b.i.d.
+ clarithromycin 500 mg b.i.d. + amoxicillin 1 g b.i.d. × 10 days.
3. Omeprazole 20 mg b.i.d.
+ clarithromycin 500 mg b.i.d. + amoxicillin 1 g b.i.d. × 10 days.
4. esomeprazole 40 mg q.d.
+ clarithromycin 500 mg b.i.d. + amoxicillin 1 g b.i.d. × 10 days.
5. Rabeprazole 20 mg b.i.d.
+ clarithromycin 500 mg b.i.d. + amoxicillin 1 g b.i.d. × 7 days.
cycline (bismuth quadruple therapy). Details regarding these
regimens can be found in Table 4. When given at the recom-
mended doses, most recent studies report intention-to-treat
(ITT) eradication rates in the range of 70–80% (121–124).
Large randomized trials suggest that the inclusion of amoxi-
cillin or metronidazole yields similar results when combined
with a PPI and clarithromycin (125). Though international
guidelines have recommended treatment durations of at least
7 days, treatment durations of 10–14 days have typically been
employed in the United States (3). A recent large trial from the
United States, which evaluated the combination of rabepra-
zole, clarithromycin, and amoxicillin, found that 7 and 10
days of therapy yielded equivalent eradication rates. The ITT
eradication rate for 7 days was 77% (95% CI 71–83%) versus
78% (95% CI 72–84%) for the 10-day regimens. This study
also reported an eradication rate of 27% for a 3-day treat-
ment regimen (123). A recent meta-analysis of seven studies
involving more than 900 patients found that a 14-day course
of clarithromycin triple therapy provided better eradication
rates than a 7-day course of therapy (Peto OR 0.62 favors
14 vs 7 days of therapy for eradication of H. pylori infec-
tion [95% CI 0.45–0.84]). There was also a trend towards
improved efficacy with 10 days of therapy compared with 7
days of therapy, which did not reach statistical significance
(126). The superiority of 14-day versus 7-day treatment du-
ration has been confirmed by a recent large randomized sin-
gle center trial from Italy (127). As a result of the falling
eradication rates with clarithromycin-based triple therapy, it
is essential to take every opportunity to optimize treatment
success. Given the results of this meta-analysis, it seems pru-
dent to recommend a 14-day course of clarithromycin triple
therapy, particularly in the United States where eradication
rates have typically been 80% or less with shorter durations
of therapy. Treatment durations of less than 7 days are clearly
associated with reduced eradication rates and are not recom-
mended. The currently available PPIs perform comparably
when used in these regimens (128, 129). Data from a recent
meta-analysis of 13 studies suggests that b.i.d. dosing of a PPI
in clarithromycin-based triple regimens is more effective than
q.d. dosing (130). Pretreatment with a PPI prior to a course
of H. pylori eradication therapy does not appear to adversely
influence treatment outcomes (131). Further, it appears that
an H
2
RA can be substituted if a patient cannot tolerate a PPI
(132).
Bismuth quadruple therapy has been advocated as a pri-
mary therapy for H. pylori (133). Bismuth quadruple ther-
apy offers eradication rates that are similar to clarithromycin
triple therapy. A recent meta-analysis including 5 random-
ized trials reported ITT and per protocol (PP) eradication
rates of 79% (95% CI 74–81%) and 85% (95% CI 81–
88%) for clarithromycin triple therapy and 80% (95% CI
77–84%) and 87% (95% CI 84–91%) for bismuth quadru-
ple therapy, respectively (134). Though this regimen has
been evaluated with an H
2
RA or PPI, a recent meta-analysis
found that quadruple therapy with a PPI provides greater
efficacy in patients with metronidazole-resistant H. pylori
strains (135). A criticism of this regimen involves its com-
plexity (q.i.d. dosing regimen and high pill count) and per-
ceived frequency of side effects. A simplified 14-day b.i.d.-
dosing regimen recently evaluated by Graham and colleagues
in the United States achieved an eradication rate of 92% (95%
CI 79–98%) (136). Another recent study reported compara-
ble eradication rates with a novel triple antibiotic capsule
given t.i.d. and a PPI b.i.d. for 10 days (137). Although mi-
nor side effects with bismuth-based quadruple therapy occur
commonly, the frequency of moderate or severe side effects
is no greater than with clarithromycin-based triple therapy
(133).
1818
Chey and Wong
It seems reasonable to consider a PPI, clarithromycin, and
amoxicillin in patients who have not previously received clar-
ithromycin and who are not allergic to penicillin. For patients
allergic to penicillin, metronidazole can be substituted for
amoxicillin. Bismuth quadruple therapy should be favored in
those allergic to penicillin or in those who have previously
been treated with a macrolide antibiotic.
Unfortunately, eradication rates yielded by clarithromycin-
based triple therapy or bismuth-based quadruple therapy are
less than 85% and may be decreasing. As such, alternative
primary therapies are necessary. Several studies from Italy
have reported eradication rates exceeding 90% with a novel
sequential therapy consisting of a PPI and amoxicillin for 5
days followed by a PPI, clarithromycin, and tinidazole for
an additional 5 days. Whether metronidazole or other imi-
dazoles can be used in place of tinidazole has not yet been
established. This regimen has achieved eradication rates su-
perior to clarithromycin-based triple therapy and was well
tolerated in children, adults, and elderly patients infected
with H. pylori (138–141). Further, sequential therapy may
be superior to clarithromycin triple therapy in patients with
clarithromycin-resistant H. pylori strains. A post hoc anal-
ysis from a large multicenter trial evaluated the efficacy of
sequential therapy versus clarithromycin triple therapy (82%
[18/22] vs 44% [7/16], P
< 0.0155) in a subset of patients
with clarithromycin-resistant H. pylori (142). In the available
studies, the reported compliance with therapy has exceeded
90% and side effects have been no greater than those expe-
rienced with clarithromycin triple therapy. Several important
questions remain to be answered regarding this promising
regimen before it can be accepted as a standard first-line ther-
apy in the United States. Perhaps most importantly, validation
of this promising new therapy in North America is necessary.
In addition, it is not clear that there is any incremental benefit
to providing antibiotic therapy sequentially rather than as a
concurrent quadruple regimen.
Predictors of H. pylori Treatment Outcome
The most important predictors of treatment failure following
anti-H. pylori therapy include poor compliance and antibiotic
resistance. There is limited evidence to suggest that smoking,
alcohol consumption, and diet may also adversely affect the
likelihood of successful eradication (143).
It is critical for clinicians to stress the importance of tak-
ing the medications as prescribed to minimize the likelihood
of treatment failure and development of antibiotic resistance.
Patients should also be informed of the most commonly re-
ported treatment-related side effects. While mild side effects
are very common with any of the recommended H. pylori
treatment regimens, significant side effects are reported in
only 5–20% (144). The most commonly reported side effects
with the PPIs include headache and diarrhea, occurring in up
to 10% of patients. To optimize their effects on gastric acid
secretion, PPIs should be taken 30–60 minutes before eating.
The most frequent side effects reported with clarithromycin
include GI upset, diarrhea, and altered taste. Common side ef-
fects associated with amoxicillin include GI upset, headache,
and diarrhea. Side effects of metronidazole tend to be dose
related and include a metallic taste in the mouth, dyspep-
sia, and a disulfiram-like reaction with alcohol consumption.
Common side effects of tetracycline include GI upset and
photosensitivity. This antibiotic should not be used in chil-
dren under 8 yr of age because of possible tooth discoloration.
Finally, bismuth compounds have been associated with dark-
ening of the tongue and stool, nausea, and GI upset (145).
Informed patients are less likely to be alarmed when side
effects that they are aware of occur and, consequently, less
likely to needlessly stop their treatment.
Antibiotic resistance must also be carefully considered
when choosing amongst the various anti-H. pylori treatment
regimens. A recent multicenter U.S. study which collected
data from 1993 to 1999 reported antibiotic resistance rates
amongst H. pylori strains of 37% for metronidazole, 10%
for clarithromycin, 3.9% for both antibiotics, and 1.4% for
amoxicillin (146). Subsequent data collected from 1998 to
2002 yielded resistance rates of 25% for metronidazole, 13%
for clarithromycin, 5% for at least 2 antibiotics, and 0.9%
for amoxicillin (147). Though these data sets are difficult to
directly compare, it appears that metronidazole and amoxi-
cillin resistance have remained relatively stable while clar-
ithromycin resistance has increased. The increasing back-
ground rate of clarithromycin resistance provides at least a
partial explanation for the decreasing efficacy of traditional
clarithromycin-containing regimens. It is quite clear that clar-
ithromycin resistance, which has been attributed to several
different point mutations in the peptidyltransferase region en-
coded in domain V of the 23S rRNA gene (142), is associated
with a high rate of treatment failure when clarithromycin-
containing regimens are employed (148–150). On the other
hand, metronidazole resistance appears to be more relative.
To some extent, metronidazole resistance can be overcome by
the use of higher doses of metronidazole and/or the addition
of a PPI to bismuth, tetracycline, and metronidazole (143).
An important study found that previous treatment with either
a macrolide or metronidazole for any reason significantly in-
creased the likelihood of H. pylori resistance to these agents
(151). As such, clinicians should routinely ask about previ-
ous macrolide or metronidazole use when deciding upon an
H. pylori treatment regimen. Further, it seems reasonable to
consider bismuth quadruple therapy with a PPI or sequential
therapy in individuals who have previously been treated with
clarithromycin or metronidazole.
Recent data suggest that bacterial and host factors also in-
fluence treatment outcomes. A systematic review and meta-
analysis including 14 studies (1,529 patients) found that
CagA-negative strains of H. pylori were associated with an
increased risk of treatment failure compared with CagA-
positive strains (risk ratio of treatment failure 2.0, 95% CI
1.6–2.4) (152). Another meta-analysis found that CYP2C19
polymorphisms, which influence the clearance of PPIs and
thus their effect on gastric acid secretion, could influence
treatment outcomes when regimens containing a PPI are used
Guideline on the Management of Helicobacter pylori Infection
1819
(153). These observations are of greatest importance to far
eastern countries where the extensive metobolizer status is
more common. Further, the inability of the clinician to readily
determine CagA status of H. pylori or the cytochrome P450
status of their patients makes these observations unlikely to
change clinical practice in the immediate future.
Salvage Therapy for Persistent H. pylori Infection
• In patients with persistent H. pylori infection, every ef-
fort should be made to avoid antibiotics that have been
previously taken by the patient.
• Bismuth-based quadruple therapy for 7–14 days is an ac-
cepted salvage therapy.
• Levofloxacin-based triple therapy for 10 days is another
option in patients with persistent infection, which requires
validation in the United States.
When faced with a patient who has failed an initial course
of therapy for H. pylori, the clinician should avoid using an-
tibiotics employed in previous treatment regimens. Because
of the expense and lack of availability, culture and antibiotic
sensitivity testing are typically not performed unless a patient
has failed at least 2 courses of therapy. Even in this circum-
stance, the usefulness of such testing is arguable as there is no
evidence to suggest that choosing a salvage regimen based
upon an understanding of the patient’s previous antibiotic
exposure is any less successful than choosing an antibiotic
regimen based upon the results of antimicrobial sensitivity
testing. Recommendations regarding salvage therapy regi-
mens are provided in Table 5.
If a patient with persistent infection has not been previ-
ously treated with clarithromycin, triple therapy with a PPI,
clarithromycin, and amoxicillin or metronidazole can be con-
sidered.
Unfortunately, most patients are initially treated with a
clarithromycin-containing regimen. In such circumstances,
the most frequently used “rescue” or “salvage” therapy is
bismuth quadruple therapy consisting of a PPI, tetracycline,
metronidazole, and bismuth (104). This salvage regimen
is widely available, inexpensive, and relatively effective. A
pooled analysis of 16 studies and 24 abstracts demonstrated
an average eradication rate of 76% (range 60–100%) for
quadruple therapy when used as second-line therapy (154).
Unfortunately, the data on quadruple therapy are difficult to
interpret as antibiotic dosing, frequency of administration,
Table 5.
Salvage Therapies for Persistent H. pylori Infection (164)
Regimen
Duration
Eradication Rates
Comments
Bismuth quadruple therapy
PPI q.d. tetracycline, Pepto Bismol, metronidazole q.i.d.
7
68% (95% CI 62–74%) Accessible, cheap but high pill count and
frequent mild side effects
Levofloxacin triple therapy
PPI, amoxicillin 1 g b.i.d., levofloxacin 500 mg q.d.
10
87% (95% CI 82–92%) Requires validation in North America
For recommendations regarding rifabutin and furazolidone, please refer to the text.
PPI
= proton pump inhibitor; q.d. = daily; q.i.d. = four times daily; b.i.d. = twice daily.
and duration of therapy vary between studies. Further, the
available studies often do not clearly report how many times
or with which antibiotics a patient has previously been treated.
As has already been discussed, disadvantages of bismuth-
based quadruple therapy include the large daily pill count
(potentially exceeding 18 pills), dosing frequency (typically
four times daily), and frequent side effects. In the hopes of
addressing some of these issues, a simplified twice-daily reg-
imen was recently evaluated and reported to yield an eradi-
cation rate of over 90% in patients who had received at least
2 previous courses of antibiotic therapy (155). Though most
international studies have utilized this regimen for 7 days
(104), a 10–14 day course is still most commonly employed
in the United States.
A number of recent studies have evaluated alternatives to
bismuth-based quadruple salvage therapy. Rifabutin, an an-
tibiotic used in the treatment of tuberculosis, has been uti-
lized as an alternative to clarithromycin in several small stud-
ies with eradication rates ranging from 38% to 91% (156–
159). In a recent study from Australia, 137 patients who had
failed therapy with omeprazole, clarithromycin, and amoxi-
cillin were treated with a 12-day course of rifabutin 150 mg,
pantoprazole 80 mg, and amoxicillin 1 g or 1.5 g daily.
The overall eradication rate was 91% and the presence of
clarithromycin or metronidazole resistance did not influ-
ence the likelihood of treatment success (160). The most
common side effects with rifabutin include rash and gas-
trointestinal complaints including nausea, vomiting, dys-
pepsia, and diarrhea. Rifabutin has been associated with
rare but potentially serious myelotoxicity and ocular toxicity
(161, 162).
Patients should be warned about the possibility of red dis-
coloration of urine while taking rifabutin.
Furazolidone, an antibiotic commonly used to treat giar-
dia, cholera, and bacterial enteritis has been evaluated as an
alternative to clarithromycin, metronidazole, or amoxicillin
for persistent H. pylori infection. Available studies utilizing
furazolidone have yielded widely variable eradication rates,
ranging from 52% to 90% (163–166). Unfortunately, fura-
zolidone is not currently marketed in the United States. Side
effects including nausea, vomiting, headache, and malaise
occur in up to a third of patients. Less frequent side effects
include hypersensitivity, hypotension, a disulfiram-like reac-
tion to alcohol, and mild, reversible hemolytic anemia (163–
167).
1820
Chey and Wong
Levofloxacin is a fluoroquinolone antibiotic with in vitro
activity against H. pylori. Levofloxacin-based triple therapy
(PPI, levofloxacin, and amoxicillin) has recently been stud-
ied as second- and third-line therapy in patients with per-
sistent H. pylori infection. In general, the available clinical
trials have involved relatively small numbers of patients and
demonstrated variable eradication rates, ranging from 63%
to 94% (168–170). A recent meta-analysis including four
randomized controlled trials found that a 10-day regimen of
levofloxacin-based triple therapy yielded superior eradication
(RR 1.41, 95% CI 1.25–1.59) and was associated with fewer
side effects (RR 0.51, 95% CI 0.34–0.75) than a 7-day course
of bismuth-based quadruple therapy. Summary eradication
rates for levofloxacin-based triple therapy and bismuth-based
quadruple therapy were 87% (95% CI 82–92%) and 68%
(95% CI 62–74%), respectively (168). A recent study from
Italy found that using rabeprazole, levofloxacin, and tinida-
zole in place of amoxicillin yielded an ITT eradication rate of
84% (171). Unfortunately, none of the studies that have eval-
uated levofloxacin-based triple therapy, have been conducted
in the United States. As such, these encouraging results re-
quire validation in the United States. The background rate
of H. pylori resistance to levofloxacin in the United States
remains largely unknown. However, preliminary data from
Canada, Italy, Belgium, and Japan suggest that such resis-
tance is found in up to 16.8% of H. pylori strains (172–175).
Whether levofloxacin resistance is absolute as is the case
with clarithromycin or more relative as with metronidazole
remains to be determined as well. While awaiting data on
antimicrobial resistance and efficacy from the United States,
given the shortage of effective, validated salvage regimens, it
seems reasonable to consider levofloxacin-based triple ther-
apy in circumstances where bismuth or clarithromycin-based
therapies are not an option.
STUDY HIGHLIGHTS
Dostları ilə paylaş: |