Articles

Helicobacter pylori eradication with a capsule containing

bismuth subcitrate potassium, metronidazole, and
tetracycline given with omeprazole versus
clarithromycin-based triple therapy: a randomised,
open-label, non-inferiority, phase 3 trial
Peter Malfertheiner, Franco Bazzoli, Jean-Charles Delchier, Krysztof Celiñski, Monique Giguère, Marc Rivière, Francis Mégraud, for the Pylera
Study Group

Summary
Background Helicobacter pylori is associated with benign and malignant diseases of the upper gastrointestinal tract, Lancet 2011; 377: 905–13
and increasing antibiotic resistance has made alternative treatments necessary. Our aim was to assess the efficacy and Published Online
safety of a new, single-capsule treatment versus the gold standard for H pylori eradication. February 22, 2011
DOI:10.1016/S0140-
6736(11)60020-2
Methods We did a randomised, open-label, non-inferiority, phase 3 trial in 39 sites in Europe, comparing the efficacy
This publication has been
and safety of 10 days of quadruple therapy with omeprazole plus a single three-in-one capsule containing bismuth corrected online.
subcitrate potassium, metronidazole, and tetracycline (quadruple therapy) versus 7 days of omeprazole, amoxicillin, The corrected version first
and clarithromycin (standard therapy) in adults with recorded H pylori infection. Patients were randomly assigned appeared at thelancet.com
on Nov 18, 2011
treatment according to a predetermined list independently generated by Quintiles Canada (Ville St-Laurent, QC,
Canada). Our study was designed as a non-inferiority trial but was powered to detect superiority. Our primary outcome See Comment page 877

was H pylori eradication, established by two negative ¹³C urea breath tests at a minimum of 28 and 56 days after the Gastroenterologie, Hepatologie
und Infektiologie, Otto-von-
end of treatment. Our assessment for non-inferiority was in the per-protocol population, with subsequent assessment Guericke-Universtität,
for superiority in the intention-to-treat population (ie, all participants randomly assigned treatment). This study is Magdeburg, Germany
registered with ClinicalTrials.gov, number NCT00669955. (Prof P Malfertheiner MD);
Department of Clinical
Medicine, University of
Findings 12 participants were lost to follow-up and 101 were excluded from the per-protocol analysis. In the per- Bologna, Bologna, Italy
protocol population (n=339), the lower bound of the CI for treatment with quadruple therapy was greater than the (Prof F Bazzoli MD); AP-HP,
pre-established non-inferiority margin of –10% (95% CI 15·1–32·3; p<0·0001). In the intention-to-treat population Groupe Hospitalier
(n=440), eradication rates were 80% (174 of 218 participants) in the quadruple therapy group versus 55% (123 of 222) Henri Mondor, Service
d’Hépatogastroentérologie et
in the standard therapy group (p<0·0001). Safety profiles for both treatments were similar; main adverse events were Centre d’Investigation
gastrointestinal and CNS disorders. Clinique 006, Créteil, France
(Prof J-C Delchier MD);
Department of
Interpretation Quadruple therapy should be considered for first-line treatment in view of the rising prevalence of
Gastroenterology, Medical
clarithromycin-resistant H pylori, especially since quadruple therapy provides superior eradication with similar safety University of Lublin, Lublin,
and tolerability to standard therapy. Poland (Prof K Celiñski MD);
Axcan Pharma Inc,
Mont-Saint-Hilaire, QC, Canada
Funding Axcan Pharma Inc.
(M Giguère PhD, M Rivière MD);
and INSERM U853, Université
Introduction Conference,4 treatment with omeprazole—a proton Victor Segalen Bordeaux 2,
Infection with Helicobacter pylori is a substantial public pump inhibitor (PPI)—and a combination of amoxicillin Bordeaux, France
(Prof F Mégraud MD)
health problem1 that affects 20–50% of people in and clarithromycin (standard therapy) is recommended
industrialised nations and up to 80% in less-developed as first-line therapy if clarithromycin resistance is 20% Correspondence to:
Prof Peter Malfertheiner,
countries.2,3 H pylori is associated with many gastro- or less. Bismuth-containing therapy is proposed in Otto-von-Guericke-Universtität,
duodenal disorders, including peptic ulcer disease, regions with high in-vitro resistance to clarithromycin Magdeburg 39120, Germany
gastric carcinoma, and gastric mucosa-associated or metronidazole, because the addition of bismuth to peter.malfertheiner@med.
ovgu.de
lymphoid tissue lymphoma.2–4 In regions with high other antibiotic regimens has been shown to improve
incidence of gastric carcinoma, eradication of H pylori is H pylori eradication.4,9
advocated to prevent the development of this disease.5–7 A previous international study,10 which assessed the
Further, patients benefit from eradication after endoscopic efficacy and safety of 10 days of omeprazole with a single
resection of early gastric carcinoma because it reduces (three-in-one) capsule containing bismuth subcitrate
the risk for metachronous gastric neoplasia.8 potassium, metronidazole, and tetracycline (quadruple
In all international guidelines,5–7 including European therapy) for H pylori eradication in patients with peptic
guidelines from the Third Maastricht Consensus ulcer disease or non-ulcer dyspepsia, reported overall

www.thelancet.com Vol 377 March 12, 2011 905

 Articles
eradication rates greater than 90%. This finding was
corroborated by a North American study,11 which
compared 10 days of this quadruple formulation with
10 days of standard therapy in patients with peptic ulcer
disease; H pylori eradication rates were 88% for quadruple
therapy and 83% for standard therapy. Quadruple therapy
eradicated 92% of in-vitro metronidazole-sensitive and
80% of metronidazole-resistant strains, whereas standard
therapy eradicated 92% of clarithromycin-sensitive
H pylori but did not eradicate the organism in 78% of
patients infected with strains resistant to clarithromycin
in vitro. Antimicrobial resistance rates have increased
since these trials,12,13 and treatments for H pylori need
assessment. Accordingly, European health authorities
have requested that quadruple therapy be compared with
standard therapy (the gold standard) for 7 days in a
clinical trial for registration purposes. We aimed to assess
efficacy and safety of quadruple therapy for 10 days versus
standard therapy given for 7 days for the eradication of
H pylori. Special attention has been directed to pre-
therapeutic H pylori resistance to clarithromycin and
metronidazole and the effect of resistance on the efficacy
of treatment.
Methods
Participants
Our trial was done at 39 sites in France, Germany,
Ireland, Italy, Poland, Spain, and the UK, between
June 11, 2008, and June 22, 2009; it was a randomised,
open-label (technicians unaware of study drug allocation),
non-inferiority, phase 3 trial. Eligible patients were aged
18 years or older with confirmed H pylori and upper
gastrointestinal symptoms. Women were eligible if they
were not pregnant or nursing, and if they were of child-
bearing potential they were required to use medically
acceptable contraception for the duration of the study
and 30 days thereafter. Patients were excluded if they had
previously used antibiotics to eradicate adequately
recorded infection with H pylori, contraindications to
study drugs, substantial organ impairment, severe or
unstable cardiopulmonary or endocrine disease,
undergone surgery of the upper gastrointestinal tract,
evidence of bleeding or iron-deficiency anaemia, Barrett’s
oesophagus or high-grade dysplasia, dysphagia, or history
of malignancy. Patients were excluded if they had history
of drug or alcohol misuse within 1 year of the trial, or
continuously used antiulcer drugs (including PPIs
during the 2 weeks before the ¹³C urea breath test),
antibiotics or bismuth compounds (more than three
times per week, 1 month before screening), systemic
glucocorticoids, non-steroidal anti-inflammatory drugs,
or anticoagulation or platelet aggregation inhibitors
(except acetylsalicylic acid ≤100 mg per day).
Written informed consent was obtained from all
patients before enrolment. Our protocol and one
amendment were approved by investigators’ research
ethics, investigational review, or independent ethics
906
boards or committees. Our study was done in accordance
with the principles of the Declaration of Helsinki and
International Conference on Harmonisation of Technical
Requirements for Registration of Pharmaceuticals for
Human Use guidelines for Good Clinical Practice.
Procedure
We screened participants up to 1 month before they were
randomly assigned treatment. Eligible patients who had
not been treated with contraindicated drugs underwent a
urea breath test during their first visit; patients who had
been treated with contraindicated drugs were allowed a
2-week washout before the breath test. All patients with
positive breath tests underwent endoscopy of the upper
gastrointestinal tract and six biopsies (four from pyloric
antrum; two from the gastric body). Antral biopsy
specimens were submitted for rapid urease test, histology,
and centralised microbiology (ie, done in one laboratory),
including culture and testing for metronidazole and
clarithromycin resistance. Patients had to be positive for
H pylori by both urea breath test and rapid urease test,
and with one confirmatory test (histology, culture, or
PCR [amendment to protocol]) for participation.
Sensitivity testing was by Etest (AB Biodisk, Solna,
Sweden). Clarithromycin resistance was defined as a
minimum inhibitory concentration (MIC) of 1·00 μg/mL
or greater, intermediate resistance was defined as
0·25 μg/mL to less than 1·00 μg/mL, and sensitivity was
defined as less than 0·25 μg/mL. Metronidazole
resistance was defined as MIC greater than 8 μg/mL and
sensitivity as 8 μg/mL or less.
Patients were randomly assigned (1:1) quadruple or
standard therapy according to a predetermined list
supplied to each site. The sponsor recruited sites and
generated the randomisation schedule through a third
party (Quintiles Canada, Ville St-Laurent, QC, Canada)
without accessing or influencing the randomisation
lists; this process also masked treatment allocation
from the investigators.
Study drugs were proprietary three-in-one capsules
containing 140 mg bismuth subcitrate potassium, 125 mg
metronidazole, and 125 mg tetracycline hydrochloride
(Axcan Pharma, Mont-Saint-Hilaire, QC, Canada);
500 mg clarithromycin capsules (Abbott, Abbott Park,
IL, USA); 500 mg amoxicillin capsules (GlaxoSmithKline,
Brentford, Middlesex, UK); and 20 mg omeprazole
capsules (AstraZeneca, LP, Wilmington, DE, USA). In
the 10-day quadruple regimen, three three-in-one
capsules were taken four-times daily (after meals and at
bedtime), and swallowed whole with 250 mL of water.
An omeprazole capsule was taken with the three single
three-in-one capsules after morning and evening meals.
In the 7-day standard regimen, one capsule of
omeprazole, two of amoxicillin, and one of clarithro-
mycin were taken twice daily (before morning and
evening meals). Patients were asked to refrain from
consuming alcohol during the entire treatment period
www.thelancet.com Vol 377 March 12, 2011
 Articles

724 individuals assessed for eligibility

284 excluded
236 did not meet inclusion criteria
18 refused to participate
30 excluded for other reasons

440 randomly assigned to treatment groups

218 assigned quadruple therapy (2 did not 222 assigned standard therapy
receive treatment)

14 did not complete the study 27 did not complete the study
5 lost to follow-up 7 lost to follow-up
2 adverse events 1 death
1 serious adverse event 4 adverse events
2 withdrew consent 1 serious adverse event
1 violation of exclusion criteria 3 withdrew consent
1 protocol violation 1 violation of inclusion criteria
2 other 3 violations of exclusion criteria
3 protocol violations
1 investigator/sponsor judgment
3 other

218 in the intention-to-treat population 222 in the intention-to-treat population

40 excluded from the per-protocol analysis 61 excluded from the per-protocol analysis
9 violations of inclusion or exclusion criteria 2 non-completers
6 compliance not 80–120% 8 violations of inclusion or exclusion criteria
4 prohibited drugs 10 compliance not 80–120%
21 missing UBTs at weeks 6 or 10 8 prohibited drugs
33 missing UBTs at weeks 6 or 10

178 in the per-protocol population 161 in the per-protocol population

Figure: Trial profile

Quadruple therapy is omeprazole, bismuth, metronidazole, and tetracycline. Standard therapy is omeprazole, amoxicillin, and clarithromycin.
UBT=¹³C urea breath test.

and for 48 h after the last dose, and not to take the study
drugs with milk, other dairy products, or antacids. Two
combination tablets (200 mg aluminium hydroxide,
200 mg magnesium hydroxide, and 25 mg simeticone;
Sanofi-Aventis, Paris, France) could be taken as rescue
medication up to four-times daily.
Our primary efficacy outcome was H pylori eradication,
established by two negative urea breath tests at least 28
and 56 days after the end of treatment (week 6 and
week 10 visits). Secondary objectives were to compare
eradication as a function of antibiotic resistance and
disease or disorder (non-ulcer dyspepsia, peptic ulcer
disease); monitor safety through adverse events,
laboratory abnormalities, and plasma bismuth concen-
trations at baseline, end of treatment, and end of study;
and assess compliance through pill count.
Patients returned for repeat examination and laboratory
studies (including measurement of plasma bismuth in
patients in the quadruple therapy group) within 1 week of
the end of trial. Patients were instructed to return all
drug containers at the end of trial visit to assess
compliance. Patients returned for the first follow-up visit
about 6 weeks after the start of treatment, but not before
28 days after the end of trial. Participants were re-
examined and reassessed with a second urea breath test
and routine laboratory studies. Breath tests were also
done on all patients who did not complete their full
course of therapy. Patients with a positive breath test at
the first follow-up underwent repeat endoscopy with
biopsies for reassessment of antibiotic resistance. If the
results of the breath test were negative, the patient was
scheduled for the end of study visit about 10 weeks after
treatment onset, when a third set of breath test, routine
examination, and clinical laboratory procedures
(including serum pregnancy testing for women of
childbearing potential) was done.

www.thelancet.com Vol 377 March 12, 2011 907

 Articles

assessed with the Cochran-Mantel-Haenszel test adjusted

OBMT (N=216) OAC (N=222)
for pooled sites (sites [within each country] or countries
Number of women 103 (48%) 100 (45%) with <12 intention-to-treat patients were incorporated
Age (years) into pooled sites). Eradication was confirmed with
Mean (SD) 48·53 (14·64) 47·95 (14·52) negative urea breath tests at both the week 6 and week 10
Median (minimum, maximum) 49·0 (18–78) 48·5 (19–81) visits. A positive breath test at week 6 suggested that
Ethnic origin treatment had failed; these patients were classified as
White 208 (96%) 216 (97%) non-eradicated. Per-protocol patients did not need
Asian 2 (1%) 1 (1%) imputation of missing test results; both week 6 and
Black or African 3 (1%) 1 (1%) week 10 assessments needed to be available for patients
Other 3 (1%) 4 (2%) whose week 6 breath test was negative. In the event of
Body-mass index (kg/m2) missing intention-to-treat assessments, non-eradicated
Mean (SD) 26·05 (3·98) 26·33 (4·38) was imputed for missing (or done outside the allocated
Median (minimum, maximum) 25·6 (19–39) 26·2 (15–43) timeframe) breath test values. All other results were
Present or past history of peptic ulcers assessed on an observed-case basis.
Yes 20 (9%) 29 (13%) The per-protocol population was used to assess primary
No 194 (90%) 189 (85%) efficacy. The primary analysis (comparing non-inferiority
¹³C urea breath test of the two drug combinations) was assessed through
Helicobacter pylori positive 215 (99·5%) 222 (100%) hypothesis testing and derivation of a two-sided 95% CI
H pylori negative 1 (0·5%) 0 with an asymptotic normal approximation procedure
Rapid urease test with continuity correction. If the lower bound of this CI
H pylori positive 215 (99·5%) 222 (100%) exceeded –10%, non-inferiority of quadruple versus
H pylori negative 1 (0·5%) 0 standard therapy could be concluded. If the lower bound
Histology* was –10% or lower non-inferiority was not supported. If
H pylori positive 211 (98%) 218 (98%) the non-inferiority of quadruple over standard therapy
H pylori negative 3 (1%) 4 (2%)
was concluded with the per-protocol population, the
Culture*
same CI was derived with the intention-to-treat
H pylori positive 181 (84%) 184 (83%)
population. If the lower bound of this CI was greater
than zero, superiority of quadruple over standard therapy
H pylori negative 31 (14%) 32 (14%)
could be concluded. If the urea breath tests were missing
Data are n (%) unless otherwise indicated. OBMT is omeprazole, bismuth, for week 6, week 10, or both, non-eradication was
metronidazole, and tetracycline. OAC is omeprazole, amoxicillin, and
assumed for this component of the analysis. Sensitivity
clarithromycin. *Percentages based on number of patients with available data.
analyses were done with the intention-to-treat population
Table 1: Baseline characteristics in the treated population without imputed data.
Two post-hoc analyses were done: the first used a
Statistical analysis different definition of eradication (one or more negative
Our hypotheses for sample size calculations were based urea breath tests at the week 6 or week 10 visit, provided
on previously documented H pylori antibiotic-resistance that it was done ≥28 days after the end of trial), and the
rates,11,14 85% quadruple therapy eradication rates and second imputed patients with missing breath tests as
80% standard therapy eradication rates, Δ (defining treatment successes. No inferential analyses were done
equivalence range or non-inferiority margin) of –10%, on safety data and no missing values were imputed. The
expected difference of 5%, power of 90%, and α of 0·025 number and percentage of patients with bismuth
(one-sided). Our trial was designed as a non-inferiority concentrations greater than 50 mg/L at all assessments
trial but was powered to detect superiority. 135 patients were reported with descriptive statistics. Bismuth
per group (completers) were needed to show non- concentrations measured at screening, end of trial, and
inferiority of the experimental drug versus active control end of study were compared with paired t tests (quadruple
on the basis of the primary efficacy variable. An estimated therapy group). This study is registered with ClinicalTrials.
30% of patients would not be included in the per-protocol gov, number NCT00669955.
analysis because of one or more major protocol deviations;
therefore, 193 patients randomly assigned treatment per Role of the funding source
group were planned to protect the trial’s power. If non- The trial was conceived and designed by the sponsor
inferiority was concluded with the per-protocol and the principal investigator (PM). The principal
population, testing for superiority with the intention-to- investigator substantially contributed to designing the
treat population was permitted (we defined the intention- trial, wrote the first and final drafts of the report and
to-treat population as all participants randomly assigned decided, in consultation with the other authors, to
treatment). SAS software (version 9.1.3 or higher) submit for publication. All authors had full access to
produced statistical outputs. Categorical variables were the data.

908 www.thelancet.com Vol 377 March 12, 2011

 Articles

Quadruple therapy Standard therapy 95% CI for difference p value

between quadruple and
standard therapy
PP population 166/178 (93%); 88·5–96·5 112/161 (70%); 61·8–76·6 15·1–32·3 <0·0001
ITT population with data imputation* 174/218 (80%); 73·9–84·9 123/222 (55%); 48·6–62·1 15·5–33·3 <0·0001
ITT population without imputed data: observed case† 174/188 (93%); 87·8–95·9 123/182 (68%); 60·3–74·3 16·7–33·3 <0·0001
Different definition in the PP population‡ 168/178 (94%); 89·9–97·3 114/161 (71%); 63·1–77·7 15·2–32·0 <0·0001
Different definition in the ITT population‡ 200/218 (92%); 87·3–95·0 152/222 (69%); 61·9–74·5 15·7–30·9 <0·0001

Data are n/N (%); 95% CI. Quadruple therapy=omeprazole, bismuth, metronidazole, and tetracycline. Standard therapy=omeprazole, amoxicillin, and clarithromycin. PP=per
protocol. ITT=intention to treat. UBT=13C urea breath test. *Since the non-inferiority of OBMT vs OAC was shown in the PP population, the same CI was derived with the ITT
population. For this analysis, if the week 6 or week 10 UBTs were not available, the patient was classified as non-eradicated. †Analysis of the ITT population without data
imputation provided similar results as those obtained with other populations. ‡We also did a post-hoc analysis with the PP and the ITT populations, whereas eradication was
defined as ≥1 negative UBT at week 6 or 10 (study days ≥39), instead of negative UBTs at both timepoints.

Table 2: Eradication rates in the PP and ITT populations

Quadruple therapy Standard therapy

Eradication rate by baseline metronidazole resistance*
Baseline metronidazole resistance Yes No Yes No
Eradication 38/42 (91%); 77·4–97·3 98/103 (95%); 89·0–98·4 28/41 (68%); 51·9–81·9 64/90 (71%); 89·0–98·4
p value 0·283 .. 0·837 ..
Eradication rate by baseline clarithromycin resistance*
Baseline clarithromycin resistance Yes No Yes No
Eradication 30/33 (91%); 75·7–98·1 106/112 (95%); 88·8–98·0 2/25 (8%); 1·0–26·0 90/106 (85%); 76·6–91·1
p value 0·426 .. <0·0001 ..
Eradication rate by baseline combined metronidazole and clarithromycin resistance*
Baseline combined metronidazole and clarithromycin resistance Yes No Yes No
Eradication 11/12 (92%); 61·5–99·8 125/133 (94%); 88·5–97·4 2/10 (20%); 2·5–55·6 90/121 (74%); 65·6–81·9
p value 0·551 .. 0·001 ..

Data ate n/N (%); 95% CI unless otherwise indicated. Quadruple therapy is omeprazole, bismuth, metronidazole, and tetracycline. Standard therapy is omeprazole, amoxicillin, and clarithromycin. PP=per protocol.
*Percentages are based on the number of patients with both baseline and post-baseline resistance data in the PP population; not all cultures provided resistance or sensitivity data.

Table 3: Eradication rates and antibiotic resistance in the PP population

Results positive breath tests) confirmed the superiority of

The figure shows the trial profile. 204 patients assigned
to receive quadruple therapy and 195 assigned to receive
standard therapy completed the study, with lack of
efficacy in 11 and 56 respectively. Two patients in the
quadruple therapy group did not receive the study
drugs; safety analyses were done on the remaining
438 patients. Table 1 shows the baseline characteristics
of the two groups.
Treatment with quadruple therapy fulfilled criteria for
non-inferiority to standard therapy in the per-protocol
population. Accordingly, the intention-to-treat population
was used for superiority testing, and quadruple therapy
was significantly better than standard therapy in eradicating
H pylori (all p<0·0001; table 2). In the per-protocol
population, eradication rates were greater for quadruple
than for standard therapy; in the intention-to-treat
population with missing urea breath test data imputed as
positive (ie, persistent infection), eradication rates were
lower for both groups (table 2). Post-hoc analysis
(eradication defined as one or more negative breath tests at
first follow-up [week 6] or end of study [week 10] and no
quadruple therapy. There was an imbalance in the number
of patients excluded from the per-protocol analysis due to
missing breath tests (figure); post-hoc analysis imputing
patients with missing breath tests as treatment successes
showed the same degree of difference between groups.
Baseline H pylori in-vitro metronidazole sensitivity was
similar between the quadruple (71% [103 of
145 participants]) and standard therapy groups (69%
[90 of 131]; p=0·695); metronidazole sensitivity did not
significantly affect the efficacy of quadruple therapy in
the per-protocol population (p=0·283; table 3). Baseline
clarithromycin sensitivity was also similar between the
quadruple (77% [112 of 145]) and standard therapy groups
(81% [106 of 131]; p=0·464); however, this sensitivity
seemed to significantly affect the efficacy of standard
therapy (p<0·0001; table 3). Simultaneous metronidazole
and clarithromycin resistance reduced efficacy only in
patients treated with standard therapy (no eradication in
eight [80%] of ten patients with resistant bacteria vs 31
[26%] of 121 with non-resistant bacteria; p=0·001).
Although few patients with peptic ulcer disease were

www.thelancet.com Vol 377 March 12, 2011 909

 Articles

(7% [16 of 222 participants]) than the quadruple therapy

Quadruple therapy Standard therapy
(N=216) (N=222) group (5% [11 of 216]). Incidence of serious TEAEs and
discontinuations due to a TEAE were similar between
TEAEs 258 223
groups (<2·0%). Four patients in the quadruple therapy
Patients with a TEAE 101 (47%) 112 (51%)
group reported a total of seven serious adverse events
Gastrointestinal disorders 65 (30%) 82 (37%)
(eczema, psychotic episode, Escherichia coli or
Dyspepsia 22 (10%) 30 (14%)
Proteus mirabilis urinary tract infection, renal artery
Diarrhoea 14 (7%) 28 (13%)
stenosis, acute renal failure, and small bowel
Abdominal pain, upper 18 (8%) 16 (7%)
neuroendocrine carcinoma); two patients in the quadruple
Nausea 14 (7%) 2 (1%)
therapy group withdrew because of treatment-limiting
Vomiting 8 (4%) 6 (3%)
adverse events (epigastric pain, dyspepsia, nausea, and
Faeces discoloured 9 (4%) 1 (1%) abdominal pain). Three patients in the standard therapy
Flatulence 3 (1%) 5 (2%) group reported a total of nine serious adverse events
Abdominal pain 3 (1%) 1 (1%) (arrhythmia, emesis, pyrexia, dehydration, under-
Eructation 3 (1%) 1 (1%) nourishment, vascular dementia, acute appendicitis,
CNS disorders 33 (15%) 29 (13%) exacerbation of chronic pancreatitis, and aggravation of
Dysgeusia 12 (6%) 22 (10%) the general condition); three patients in the standard
Headache 18 (8%) 7 (3%) therapy group withdrew because of treatment-limiting
Dizziness 4 (2%) 1 (1%) adverse events (epigastric pain, tachycardia, hyperhidrosis,
Somnolence 3 (1%) 0 and cardiovascular complications). The sole death during
Infections and infestations 17 (8%) 18 (8%) the trial was suicide of a patient in the standard therapy
Nasopharyngitis 6 (3%) 7 (3%) group 2 months after study entry—we did not deem this
Upper respiratory tract infection 3 (1%) 2 (1%) related to the study drug. No clinically significant changes
Influenza 0 4 (2%) were reported in physical examination, vital signs, or
General disorders and administration-site conditions 14 (7%) 7 (3%) laboratory indices. Bismuth plasma concentrations
Malaise 5 (2%) 3 (1%) increased slightly with quadruple therapy (45 [22%] of
Pyrexia 3 (1%) 1 (1%) 209 patients in this group developed plasma bismuth
Respiratory, thoracic, and mediastinal disorders 8 (4%) 3 (1%) concentrations of 4–20 μg/L after having undetectable
Cough 4 (2%) 1 (1%) baseline values of <4 μg/L). All bismuth concentrations
Musculoskeletal and connective tissue disorders 7 (3%) 3 (1%) were below the toxic threshold (50 μg/L).15 Five (3%) of
Back pain 4 (2%) 1 (1%) 200 patients treated with quadruple therapy had detectable
Psychiatric disorders 3 (1%) 5 (2%) plasma bismuth concentrations at the end of study visit
Insomnia 0 3 (1%) versus three (1%) of 210 who had detectable plasma
bismuth concentrations at baseline.
Quadruple therapy is omeprazole, bismuth, metronidazole, and tetracycline. Standard therapy is omeprazole,
amoxicillin, and clarithromycin. TEAE=treatment-emergent adverse event.
Discussion
Table 4: TEAEs in the treated population Our results show that 10 days of treatment with quadruple
therapy yields H pylori eradication rates superior to 7 days
included in this trial (table 1), efficacy was still assessed of treatment with standard therapy. Standard therapy is
in this group. Quadruple (93% [147 of 158]) and standard at present the standard first-line treatment in Europe and
therapy (67% [95 of 141]; p<0·0001) efficacy in patients elsewhere,4–7 but failure rates in about 40% of patients
with non-ulcer dyspepsia was similar to the overall study have been reported and continue to increase.14,16
population. Patients with peptic ulcer disease responded Antibiotic resistance is the main factor that contributes
better to quadruple (95% [18 of 19]) than to standard to the failure of PPI-based triple therapy to adequately
therapy (83% [15 of 18]; p=0·340); however, this result eradicate H pylori. Resistance-inducing mutations happen
should be interpreted with caution in view of the small spontaneously during bacterial replication, and previous
number of patients in this subgroup. Compliance antibiotic treatment for other infections favour the
exceeded 95% in both treatment groups. selection of drug-resistant H pylori;17 resistance can rapidly
The proportion of patients with treatment-emergent emerge during the first course of treatment18 and parallels
adverse events (TEAEs) were similar between the national antibiotic consumption.19 Clarithromycin is a key
treatment groups (table 4). Gastrointestinal symptoms, component of many first-line PPI-based triple therapies,
including dyspepsia and diarrhoea, were the most and resistance to it is constantly rising (paralleled by
commonly reported TEAEs and were more common in increasing standard therapy failure).20 Conversely, the
the standard than the quadruple therapy group. efficacy of metronidazole-containing combination therapy
Severe TEAEs (dyspepsia, upper-abdominal pain, (with bismuth and tetracycline) is maintained in patients
nausea, eructation, headache, dysgeusia, and general- harbouring metronidazole-resistant H pylori.9,21 The
ised infection) were more common in the standard discrepancy is explained partly by metronidazole resistance

910 www.thelancet.com Vol 377 March 12, 2011


as assessed in vitro, but does not adequately relate to
in-vivo states, partly because of metronidazole–bismuth
synergism that overcomes resistance.22 Further,
metronidazole resistance is complex and associated with
the presence of redox-related bacterial-gene mutations
(rdxA, frxA, ferric uptake regulator protein).23
Bismuth has an established history in the treatment
of H pylori.24 Colloidal bismuth subcitrate has potent
anti-H pylori activity (MIC 4–32 μg/mL), and in-vitro
resistance has not been induced. Further, bismuth
increases eradication when included in double, triple,
and quadruple regimens.22,25 In our trial, eradication
rates with quadruple therapy were similar
(about 88–93%) to those of previous studies that used
the same three-in-one capsule with 20 mg omeprazole
twice daily.10,11 Such high eradication rates contrast with
those reported from a meta-analysis by Luther and
colleagues26 that compared the efficacy of bismuth-
based quadruple therapy (about 78% eradication) with
standard therapy (about 77%). Although standard
therapy dosing regimens were consistent across trials,
they were highly variable for bismuth-based quadruple
therapy; further, choice of bismuth salt also varied.10,11,26
Although we could postulate that inclusion in the three-
in-one capsule is ideal, head-to-head trials are needed
to establish the bismuth dosing regimen that provides
maximum eradication.
Concerns have been raised about toxic effects related
to bismuth. There have been rare reports of heavy-metal
poisoning and encephalopathy at plasma bismuth
concentrations greater than 50 μg/L in patients treated
with bismuth subgallate and greater than 20 g per day
bismuth subnitrate for 6 months to 20 years.27–29 In our
trial, with little bismuth exposure and use of a colloidal
bismuth subcitrate formulation, bismuth concentrations
were below the toxic threshold. Data from a recent
systematic review and meta-analysis of 35 randomised
controlled trials showed that stool discolouration
(clinically irrelevant) was the only adverse effect
significantly associated with bismuth ingestion.24
Inclusion of antisecretory therapy (PPIs) in H pylori-
eradication regimens is crucial to optimise the local
activity of antibiotics via synergism,30 and our trial
clarifies the timing of PPI in relation to meals. By contrast
with giving a PPI with amoxicillin and clarithromycin
(whose stability and activity are highly dependent on
high pH), there is no such need with tetracycline
and metronidazole.
In our trial, two negative urea breath tests in the allocated
timeframe were needed for a patient to be deemed
successfully treated. In view of the high accuracy of the
breath test, treatment guidelines and most studies and
meta-analyses judge eradication to be confirmed after one
breath test 4–6 weeks after completion of treatment.4,26,31,32
In our study, seven patients had a negative breath test at
week 6 and a positive test at week 10 (two on quadruple
therapy, five on standard therapy); confirmation of a
www.thelancet.com Vol 377 March 12, 2011
Articles
negative test was obtained in 437 of 438 patients. The
extremely conservative approach of imputing all patients
with one breath test outside the allocated timeframe as
non-eradicated explains the underestimation of eradication
in the intention-to-treat population.
Criticisms of our study might include the varying
treatment duration in each group and choice of
metronidazole-resistance test. The effect of varying
standard therapy durations has been assessed in a recent
meta-analysis of 21 studies, with data showing that the
extension of standard therapy beyond 7 days provided a
slight increase (about 5%) in eradication when the longest
duration (14 days) was assessed.31 However, when the four
highest-quality trials (as assessed by the Jadad scale)33 from
this analysis were considered, prolonging treatment with
standard therapy beyond 7 days (with extension up to
10 days) was not associated with additional benefit.31 A
randomised trial of patients infected with H pylori34 showed
that extending treatment with standard therapy for up to
Panel: Research in context
Systematic review
We searched published work with PubMed, Ovid Medline,
Abridged Index Medicus, and various gastrointestinal
congresses (including the American Gastroenterological
Association, Digestive Disease Week, and the United European
Gastroenterology Week). Publications were limited to 1980 or
later. Search terms included “Helicobacter pylori”, “peptic ulcer
disease”, “eradication”, “guideline”, “therapy”, and
“treatments”. Authors independently extracted and validated
references. Data from systematic reviews of randomised
controlled trials, individual randomised controlled trials,
systematic reviews of cohort studies, cohort studies,
systematic reviews of case-control studies, case-control
studies, case series, and opinions without explicit critical
appraisal were prioritised for inclusion in descending order.
Our search was limited to studies in English.
Interpretation
Our trial compared the efficacy and safety of omeprazole,
amoxicillin, and clarithromycin (standard therapy; first-line
treatment in international guidelines) and omeprazole,
bismuth subcitrate potassium, metronidazole, and
tetracycline (quadruple therapy) and, as shown by the
results of our systematic review, is the first large-scale
comparative study of these treatments for the eradication
of H pylori in nearly a decade. Data from our study
corroborate the safety and tolerability of bismuth in Europe
and highlight the effect of antibiotic resistance on
eradication efficacy between combination treatments.
Clarithromycin-resistant H pylori continues to increase in
prevalence, and although rates of metronidazole resistance
are also high, the results from our study showed that
clarithromycin resistance reduces the efficacy of standard
therapy, whereas resistance to metronidazole has a slight
effect on the efficacy of quadruple therapy.
911
 Articles
14 days did not provide any benefit over 7 days of treatment.
Regarding quadruple therapy, 10 days of treatment is
advocated by international guidelines5–7 and is the approved
duration of treatment in some countries. 7 days of
quadruple therapy has not been rigorously assessed.
There is no recommendation for metronidazole testing.
Etest is convenient to use and its results relate well with
agar dilution;35 although Etest can overestimate resistance,
our results are consistent with previous data.36
Our trial is the first large-scale comparative study of the
efficacy and safety of standard and quadruple therapy for
the eradication of H pylori in nearly a decade (panel). In
view of the variable availability of bismuth across Europe,
the findings of our trial provide assurance to clinicians
on the safety and tolerability of bismuth. Furthermore,
we highlight the effect of antibiotic resistance on
eradication efficacy between combination therapies.
Although several other new eradication strategies have
been proposed, their efficacy and use in practice need to
be corroborated.6,20,37–39
10 days of treatment with quadruple therapy was
superior to 7 days of standard therapy for H pylori
eradication in patients with or without the presence or
history of peptic ulcer disease. Clarithromycin-resistant
H pylori isolates continue to increase in prevalence
throughout the world, and although rates of metronidazole
resistance are also high, clarithromycin resistance
reduces the efficacy of standard therapy whereas
metronidazole resistance has little effect on the efficacy
of quadruple therapy. In regions with high levels of
clarithromycin resistance, treatment with quadruple
therapy should be considered as first-line therapy for
H pylori eradication.
Contributors
PM, MG, and MR contributed to the study concept and design.
PM, FB, J-CD, and KC enrolled and treated patients and collected
data. FM provided central microbiological review. All authors
contributed to analyses and interpretations of results and the writing
or review of the paper.
Pylera Study Group investigators
France (30 patients) A Aisene, M Bougnol, J Cassigneul, P Coulom,
P Houcke, D Lamarque, D Lamouliatte, J-F Roques, A Thevenin,
J-C Delchier, S Ecuer, G Bagnouls, T Helbert. Germany (182) H Andree,
H Bouzo, W Brandt, H-J Cordes, A Dettmer, H Juergens,
P Malfertheiner, I Rehmann, A Ryschka, T Schaefer. Ireland (3)
C O’Morain. Italy (18) F Bazzoli, G Gasbarrini, L Cipolletta,
V Stanghellini. Netherlands (0) E Kuipers, W A de Boer. Poland (184)
K Celiñski, W Gachowski, J Huk, Z Jamrozik-Kruk, W Karnafel,
D Kleczkowski, K Kujawski, K Linke, R Petryka, D Serwin, B Wozniak,
A Wysokinski. Spain (7) M Castro, J L Calleja, X Calvet, J Gisbert,
E Domínguez Muñoz. UK (7) C Bundy, I Orpen, D St John Mohr,
I Parker, T Cahill.
Conflicts of interest
PM, FB, J-CD, KC, and FM have received payment from Axcan Pharma for
research and clinical trials. MG and MR are employees of Axcan Pharma.
Acknowledgments
The authors thank Jonathan M Wert for editorial assistance in the
preparation of this report, which was supported by Axcan Pharma.
References
1 Moayyedi P, Hunt RH. Helicobacter pylori public health
implications. Helicobacter 2004; 9 (suppl 1): 67–72.
912
2 Suerbaum S, Michetti P. Helicobacter pylori infection. N Engl J Med
2002; 347: 1175–86.
3 Malfertheiner P, Chan FK, McColl KE. Peptic ulcer disease. Lancet
2009; 374: 1449–61.
4 Malfertheiner P, Megraud F, O’Morain C. Current concepts in
the management of Helicobacter pylori infection: the Maastricht III
Consensus Report. Gut 2007; 56: 772–81.
5 Asaka M, Kato M, Takahashi S, et al, for the Japanese Society
for Helicobacter Research. Guidelines for the management
of Helicobacter pylori infection in Japan: 2009 revised edition.
Helicobacter 2010; 15: 1–20.
6 Chey WD, Wong BCY. American College of Gastroenterology
guideline on the management of Helicobacter pylori infection.
Am J Gastroenterol 2007; 102: 1808–25.
7 Fock KM, Katelaris P, Sugano K, et al, for the Second Asia-Pacific
Conference. Second Asia-Pacific Consensus Guidelines for
Helicobacter pylori infection. J Gastroenterol Hepatol 2009;
24: 1587–1600.
8 Fukase K, Kato M, Kikuchi S, et al, for the Japan Gast Study Group.
Effect of eradication of Helicobacter pylori on incidence of
metachronous gastric carcinoma after endoscopic resection of early
gastric cancer: an open-label, randomised controlled trial. Lancet
2008; 372: 392–97.
9 Graham DY, Opekun AR, Belson G, El-Zimaity HM, Carlson MR.
Novel bismuth-metronidazole-tetracycline triple-layer tablet for
treatment of Helicobacter pylori. Aliment Pharmacol Ther 2005;
21: 165–68.
10 O’Morain CA, Borody T, Farley A, et al, for the International
Multicentre Study. Efficacy and safety of single triple capsule of
bismuth biskalcitrate, metronidazole, and tetracycline given with
omeprazole for eradication of Helicobacter pylori: an international
multicenter study. Aliment Pharmacol Ther 2003; 17: 415–20.
11 Laine L, Hunt R, El-Zimaity H, Nguyen B, Osato M, Spénard J.
Bismuth-based quadruple therapy using a single capsule of
bismuth biskalcitrate, metronidazole, and tetracycline given with
omeprazole versus omeprazole, amoxicillin, and clarithromycin
for eradication of Helicobacter pylori in duodenal ulcer patients:
a prospective, randomized, multicenter, North American trial.
Am J Gastroenterol 2003; 98: 562–67.
12 De Francesco V, Zullo A, Ierardi E, et al. Phenotypic and genotypic
Helicobacter pylori clarithromycin resistance and therapeutic
outcome: benefits and limits. J Antimicrob Chemother 2010;
65: 327–32.
13 Graham DY, Fischbach L. Helicobacter pylori treatment in the era
of increasing antibiotic resistance. Gut 2010; 59: 1143–53.
14 Mégraud F. H pylori antibiotic resistance: prevalence, importance,
and advances in testing. Gut 2004; 53: 1374–84.
15 Hillemand P, Pallière M, Laquais B, Bouvet P. Bismuth treatment
and blood bismuth levels. Sem Hop 1977; 53: 1663–69 (in French).
16 De Francesco V, Margiotta M, Zullo A, et al. Prevalence of primary
clarithromycin resistance in Helicobacter pylori strains over a 15 year
period in Italy. J Antimicrob Chemother 2007; 59: 783–85.
17 Blaser MJ. The genetic gymnastics of our indigenous microbes.
N Engl J Med 2002; 346: 2083–85.
18 Koletzko S, Richy F, Bontems P, et al. Prospective multicentre study
on antibiotic resistance of Helicobacter pylori strains obtained from
children living in Europe. Gut 2006; 55: 1711–16.
19 Boyanova L. Prevalence of multidrug-resistant Helicobacter pylori
in Bulgaria. J Med Microbiol 2009; 58: 930–35.
20 Mégraud F, Lehours P. Helicobacter pylori detection and
antimicrobial susceptibility testing. Clin Microbiol Rev 2007;
20: 280–322.
21 Graham DY, Osato MS, Hoffman J, et al. Metronidazole containing
quadruple therapy for infection with metronidazole resistant
Helicobacter pylori: a prospective study. Aliment Pharmacol Ther 2000;
14: 745–50.
22 Goodwin CS, Marshall BJ, Blincow ED, Wilson DH, Blackbourn S,
Phillips M. Prevention of nitroimidazole resistance in
Campylobacter pylori by coadministration of colloidal bismuth
subcitrate: clinical and in vitro studies. J Clin Pathol 1988;
41: 207–10.
23 Tsugawa H, Suzuki H, Satoh K, et al. Two amino acids mutation
of ferric uptake regulator determines Helicobacter pylori resistance
to metronidazole. Antioxid Redox Signal 2011; 14: 15–23.
www.thelancet.com Vol 377 March 12, 2011

24 Ford AC, Malfertheiner P, Giguere M, Santana J, Khan M,
Moayyedi P. Adverse events with bismuth salts for Helicobacter
pylori eradication: systematic review and meta-analysis.
World J Gastroenterol 2008; 14: 7361–70.
25 Chiba N. Effects of in vitro antibiotic resistance on treatment:
bismuth-containing regimens. Can J Gastroenterol 2000; 14: 885–89.
26 Luther J, Higgins PD, Schoenfeld PS, Moayyedi P, Vakil N,
Chey WD. Empiric quadruple vs triple therapy for primary
treatment of Helicobacter pylori infection: systematic review and
meta-analysis of efficacy and tolerability. Am J Gastroenterol 2010;
105: 65–73.
27 Bader JP. The safety profile of De-Nol. Digestion 1987;
37 (suppl 2): 53–59.
28 Buge A, Rancurel G, Poisson M, et al. 20 cases of acute
encephalopathy with myoclonia during treatment with oral bismuth
salts. Ann Med Intern 1974; 124: 877–88.
29 Slikkerveer A, de Wolff FA. Pharmacokinetics and toxicity of
bismuth compounds. Med Toxicol Adverse Drug Exp 1989; 4: 303–23.
30 Mégraud F. Update on therapeutic options for Helicobacter
pylori-related diseases. Curr Infect Dis Rep 2005; 7: 115–20.
31 Fuccio L, Minardi ME, Zagari RM, Grilli D, Magrini N, Bazzoli F.
Meta-analysis: duration of first-line proton-pump inhibitor based
triple therapy for Helicobacter pylori eradication. Ann Intern Med
2007; 147: 553–62.
32 Saad RJ, Schoenfeld P, Kim HM, Chey WD. Levofloxacin-based
triple therapy versus bismuth-based quadruple therapy for
persistent Helicobacter pylori infection: a meta-analysis.
Am J Gastroenterol 2006; 101: 488–96.
www.thelancet.com Vol 377 March 12, 2011
Articles
33 Jadad AR, Moore RA, Carroll D, et al. Assessing the quality
of reports of randomized clinical trials: is blinding necessary?
Control Clin Trials 1996; 17: 1–12.
34 Zagari RM, Bianchi-Porro G, Fiocca R, Gasbarrini G, Roda E,
Bazzoli F. Comparison of 1 and 2 weeks of omeprazole, amoxicillin
and clarithromycin treatment for Helicobacter pylori eradication:
the HYPER study. Gut 2007; 56: 475–79.
35 Glupczynski Y, Broutet N, Cantagrel A, et al. Comparison of the
E test and agar dilution method for antimicrobial susceptibility
testing of Helicobacter pylori. Eur J Clin Microbiol Infect Dis 2002;
21: 549–52.
36 Fischbach L, Evans EL. Meta-analysis: the effect of antibiotic
resistance status on the efficacy of triple and quadruple first-line
therapies for Helicobacter pylori. Aliment Pharmacol Ther 2007;
26: 343–57.
37 Essa AS, Kramer JR, Graham DY, Treiber G. Meta-analysis:
four-drug, three antibiotic, non-bismuth-containing “concomitant
therapy” versus triple therapy for Helicobacter pylori eradication.
Helicobacter 2009; 14: 109–18.
38 Graham DY, Shiotani A. New concepts of resistance
in the treatment of Helicobacter pylori infections.
Nat Clin Pract Gastroenterol Hepatol 2008; 5: 321–31.
39 Gisbert JP, Calvet X, O’Connor A, Mégraud F, O’Morain CA.
Sequential therapy for Helicobacter pylori eradication: a critical
review. J Clin Gastroenterol 2010; 44: 313–25.
913