Antibiotic resistance is the most important factor responsible for the declining success rate of H. pylori eradication therapy. Surveillance of H. pylori antibiotic resistance is mandatory in order to adapt the antibiotic combination to local resistance patterns.
The rates of H. pylori resistance to amoxicillin (0%) and metronidazole (85%) observed in Senegal were in keeping with those reported in this country in 1999–2000 (0% and 90%, respectively). In contrast, resistance to levofloxacin rose from 0% in 1999–2000 to 15% in 2007–2008, mirroring the situation in several European and Asian countries, with increases from 3% in 1999 to 15% in 2004 in France, from 11% in 2003 to 22% in 2005 in Germany, from 3% in 1998 to 12% in 2003 in Taiwan, and from 0% in 1987 to 33% in 2003 in South Korea[15–19]. This probably reflects the increasing use of quinolones in these countries[20, 21], highlighting the importance of appropriate use of this class of antibiotics to limit the development of antimicrobial resistance. Interestingly, a significant increase in ciprofloxacin resistance has also been observed in Escherichia coli isolates from Senegalese patients with community-acquired urinary tract infections (10% in 2004, 22% in 2006).
The lack of resistance to amoxicillin and tetracycline in this study indicates that H. pylori resistance to these agents is probably exceptional (0 to 1.3%) and low (0 to 4.4%), respectively, whatever the continent[21, 23]. The high prevalence of resistance to these antibiotics observed in some developing and emerging countries probably reflects inadequate laboratory facilities[5–7, 9] (e.g. 6% in Ethiopia, 6.6% in Bangladesh, 39.0% in Brazil, 59% in Iran for resistance to amoxicillin; 9.0% in Brazil, 15.0% in Bangladesh, 27.0% in Chile for resistance to tetracyclin). In our study, only one strain was resistant to clarithromycin (MIC>256 mg/l), likely reflecting low macrolide consumption, although no data on antibiotic consumption for clarithromycin but also erythromycin and azithromycin are available for Senegal. This resistance was due to a point mutation at nucleotide position 2143 (A2143G) of the 23S rRNA gene, a well-known mutation described worldwide. Resistance to clarithromycin is increasing in most countries in Central, Western and Southern Europe, as well as in East Asia, and has now reached more than 20% in these areas. Clarithromycin-containing regimens should thus be recommended for first-line empirical H. pylori eradication therapy.
The prevalence of resistance to levofloxacin was 15% in Senegal. In the 16 levofloxacin-resistant strains, N87I and D91N were the most common mutations in the QRDR region of gyrA. The major mutation (N87I) has been detected sporadically worldwide, indicating that local founder effects have resulted in local propagation of resistant clones. D91N is a well-known mutation detected worldwide. As three strains with levofloxacin MICs of 8 mg/l displayed no mutations in the QRDR of gyrA, we sequenced the entire gyrA and gyrB genes of these strains but found no mutations. This points to the presence of another mechanism such as reduced drug accumulation. The Maastricht IV/Florence Consensus Report recommends that empirical use of levofloxacin should be abandoned when the prevalence of resistance reaches 15–20%. Levofloxacin-containing therapy should therefore only be used in Senegal if compatible with the antimicrobial resistance pattern of the strain.
The worldwide prevalence of resistance to metronidazole ranges from 20–40% in Europe and the USA, to 50–80% in developing countries, Iran, India and Egypt displaying the highest rates of resistance (80 to 100%)[6, 21, 24, 25]. Metronidazole is used extensively to treat parasitic diseases in tropical countries, which probably explains the higher prevalence of resistance to this drug. Although standard metronidazole susceptibility testing lacks reproducibility, trends of low, medium, or high prevalence rates observed at a population level seem real. Then, metronidazole-containing triple therapy should not be recommended for first-line empirical H. pylori eradication therapy. It should be used within bismuth-based quadruple regimen or sequential treatment for 14 days.