Antimicrobial resistance of Helicobacter pylori
Elzbieta Rozynek, Danuta Dzierzanowska, Danuta Celinska-Cedro,
Anna Gzyl and Janusz Jeljaszewicz
Children's Memorial Health Institute, National Institute of Hygiene, Warsaw, Poland
Helicobacter pylori infections are typically treated with
metronidazole, a second antibiotic and bismuth salts or omeprazole. In patients with metronidazole-susceptible
strains, this therapy can result in cure rates of greater than 90% (1,2). However, the efficacy of treatment is
limited by primary drug resistance or rapidly-emerging drug resistance during treatment (3,4,5,6). In European
countries, resistance rates of H. pylori to metronidazole before treatment range from 7 - 50% (4,7). It has been
suggested, however, that modest rates of in vitro metronidazole resistance may not correlate with poor cure when a multi-drug
treatment regime is used (8). The mechanism of metronidazole resistance in H.
pylori is thought to be associated with a lack of reduction
of the drug's NO2
group (9; see Hoffman and Berg, page 3).
Macrolides are newer therapeutic agents used in treating H.
pylori infections. Among them, erythromycin is rather ineffective
because it is acid-labile; clarithromycin is more acid-stable and clinically more effective (10). When resistance
to clarithromycin does present itself, it may be due to previous exposure to erythromycin (11). Recent data indicates
that primary resistance of H. pylori
isolates to clarithromycin and roxithromycin in Europe varies from 5 -15% (7, 11) and may be an important factor
determining the cure rate (11). Investigations have shown that clarithromycin-resistance is linked to mutations
in the 23S rRNA gene (12), while resistance to quinolones is a result of alterations in the
gyr A gene (13).
The most frequent H. pylori
infections in children have been associated with chronic active gastritis (14). This chronicity suggests that it
is important to examine primary antibiotic resistance through susceptibility testing to select appropriate agents
for the eradication of H. pylori.
The aim of our study was to determine primary resistance of H.
pylori to metronidazole and other antibiotics in isolates
from children with gastroduodenal disease. In addition, we assessed the impact of primary metronidazole resistance
on the rate of eradication of H. pylori
a triple therapy regimen: metronidazole, amoxicillin and ventrisol (bismuth salts).
Between 1993 and 1996, 130 strains of H. pylori were collected before treatment. Isolates were cultured from antral biopsies
and were identified using standard methods (7). Different methods have been developed to determine antibiotic and
include the disk diffusion method, agar dilution, E-test and the miniwell format method with egg yolk agar (15-18).
We determined resistance using E-testing to obtain minimum inhibitory concentrations (MICs) and used the following
resistance thresholds (11,15,17): amoxicillin (>8 µg/ml), tetracycline (>8 µg/ml), erythromycin
(>8 µg/ml), clarithromycin (>8 µg/ml), ciprofloxacin (>8 µg/ml) and metronidazole (>8
Figures 1 and 2 show the distributions of MICs we obtained for metronidazole and macrolides, respectively, among
the 130 isolated strains. Table 1 lists the susceptibility rates to metronidazole and five other antimicrobials.
Over 50% of the tested strains showed high level resistance to metronidazole (MIC90 > 32 µg/ml). Nine
isolates showed moderate susceptibility (MIC90 = 8 µg/ml) to the drug. Clarithromycin was more active (MIC90
= 8 µg/ml) than erythromycin (MIC90 = 64 µg/ml). None of the 130 strains tested were resistant to amoxicillin
or tetracycline, while one strain showed resistance to ciprofloxacin. Over 16% of pre-treatment H. pylori strains
isolated from the children were simultaneously resistant to metronidazole, erythromycin, and clarithromycin (Table
2). Such multiple resistances may predispose patients to treatment failure.
To study the impact of primary and emerging resistance rates on the rate of H.pylori eradication, we also evaluated the isolation frequency of metronidazole-resistant
strains of H. pylori
in 50 children (35 with gastritis and 15 with duodenal ulcers) before and after they received specific treatment
(20 mg/kg/ per day of metronidazole and 40 mg/kg per day of amoxicillin for 10 days, with the addition of bismuth
salts for 6 weeks). Eradication rates were determined by histologic, culture and PCR methods 4-6 weeks after completion
of therapy. Metronidazole-resistant strains isolated before and after specific treatment were typed using the PCR-RFLP
method with the H. pylori
NCTC 11637 probe.
Prior to treatment (Table 3), 40% of the pre-treatment strains demonstrated a high level of primary resistance
(MIC=16 - >32 µg/ml) to metronidazole. After 6 weeks of triple therapy, 58% of those children failed to
fully respond to treatment (Table 4). Highly resistant strains of H.
pylori isolated from these children before and after treatment
had identical restriction patterns. This low success rate for eradication may be explained by the high rate of
primary resistance to metronidazole.
Glupczynski has described the European prevalence of resistant strains to metronidazole as <10-50%, ciprofloxacin
<1%, and macrolides 5-15%. (5). Our study showed high primary resistance (51%; 67/130) to metronidazole among
tested H. pylori
strains, 16% (22/130) resistance to macrolide antibiotics, and 0.8% (1/130) to ciprofloxacin. Resistance to metronidazole
described by other authors in Poland has ranged from 50 to 53%. [19-21]. H.
pylori resistance to metronidazole and clarithromycin has
also reported to be very high (50%) in developing countries such as Peru. (18). The high percentage of H. pylori resistance
to clarithromycin in Poland is associated with the high rate of prescription of macrolide antibiotics to children.
It should be noted that a high percentage (16.9%) of strains was simultaneously resistant to metronidazole and
macrolides. Zia et al showed no statistical difference in the prevalence of metronidazole resistance between clarithromycin-sensitive
isolates (37%) and clarithromycin-resistant isolates (55%); these authors observed cross resistance among macrolides.
These findings and the clinically relevant correlation between high resistance (40%) to metronidazole before treatment
and low rates of eradication of H. pylori (62%) in 50 children with chronic active gastritis underscores the need for
judicious and appropriate use of antibiotics.
- Seppälä K, Färkkilä M, Nuutinen
H, et al. Scand J Gastroenterol 1992; 27: 973-976.
- Xia HX, Daw MA, Sant S, et al. Eur J Gastroenterol
and Hepatol 1993(b); 5:141-144.
- Rautelin H. Seppälä K, Renkonen O, et al.
Antimicrob Agents Chemother 1992; 36:163-166.
- Noach LA, Langenberg WL, Bertola MA, et al. Scand J
Infect Dis 1994; 26:321-327.
- Glupczynski Y, Burette A. Lancet 1992; 339:54-55 (Letter)
- Celi ska-Cedro D, Teisseyre M, Woynarowski M, et al.
Gastroenterologia Polska 1995; 2: 115-119.
- Lee A, Megraud F. Helicobacter pylori techniques for
clinical diagnosis basis research in Lee A, Megraud F, eds.1996, 17-28 W.B. Saunders Company Ltd, London.
- Graham DY, de Boer WA, Tytgat NJ. Am J. Gastroenterol
- Cederbrant G, Kahlmeter G, Ljung A: J Antimicrob Chemother
- Graham DY, Operkun AR, Klein PD. J Clin Gastroenterol
- Xia HX, Buckley M, Keane CT, et al. J Antimicrob Chemother
- Stone GG, Shortridge D, Versalovic J, et al. Antimicrob
Agents Chemother 1997; 41:712-714.
- Moore RA, Beckthold B, Wong S, et al. Antimicrob Agents
Chemother 1995; 39:107-111.
- Glassman MS. Clin Pediatrics 1992; 8:481-485.
- Xia H, Keane CT, Beattie S,m et al. Antimicrob Agents
Chemother 1994; 38:2357-2361.
- Cederbrant G, Kahlmeter G, Ljungh A, et al. J Antimicrob
Chemother 1993; 31:65-71.
- Glupczynski Y, Labbe M, Hansen W, et al. J Clin Microbiol
1991; 29: 2072-2075.
- Vasquez A, Valdez Y, Gilman RH, et al. J Clin Microbiol
- Andrzejewska E, Klincewicz H. Gastroenterolgia Polska
- Gosciniak G, Glupczynski Y, Butzler JP. J Physiol Pharmacol
1995; 46(suppl2):31 (abstract).
- Rozynek E, Dzierzanowska D, Celinska-Cedro D, et al.
Gastroenterolgia Polska 1995; 2:129-133.