- Email: [email protected]
The predictive value of Helicobacter pylori in-vitro metronidazole resistance
CORRESPONDENCE The predictive value of Helicobacter pylori in-vitro metronidazole resistance Resistance to antibiotics and the rate of therapeutic failure with Helicobacter pylori infection are increasing worldwide. In-vitro sensitivity tests may help the clinician to establish the correct therapeutic approach, but in-vitro determination of resistance to metronidazole poses particular problems for clinical microbiologists that may be related to bacterial characteristics or to pitfalls in the technical procedures used to test sensitivity [1,2]. Henriksen et al. [3] recently highlighted the influence that the laboratory handling of H. pylori may have on the results of in-vitro determination of metronidazole resistance, in that bacterial storage significantly reduced the rate of metronidazole resistance, and that isolates retaining in-vitro resistance had significantly higher rates of treatment failure than isolates that lost resistance after storage for 3 years. The results of this study underlined some important questions regarding microbiological techniques, and indicated the need to perform experimental studies designed specifically to help the clinical microbiologist to improve laboratory results that may guide antimicrobial therapy. We have two specific comments on the conclusions reached by Henriksen et al. [3]. The first relates to the possible explanations for the differences observed in metronidazole susceptibility results after storage. It has been shown by several groups [4,5] that H. pylori generally comprises a population heterogeneous for genetic traits and antibiotic resistance. To investigate the heterogeneity of H. pylori cultured from gastric biopsies, we recently used a single-colony approach, based on immediately subculturing bacterial isolates from each patient to yield 15–20 single colonies. The phenotypic and genotypic characteristics of each colony were then studied in parallel. The results showed that a proportion of the patients carried mixed populations of antibiotic-sensitive and antibiotic-resistant microorganisms [6]. More interestingly, the distribution of these subpopulations was highly variable within patients, with metronidazole-resistant colonies accounting for 40–90% of the entire bacterial population. It was also demonstrated that metronidazole resistance was associated significantly with an intact cag pathogenicity
10.1111/j.1469-0691.2004.00985.x
island, although the composition of this gene cluster did not affect resistance to clarithromycin or amoxycillin [6]. Based on these observations, it can be hypothesised that the differences in susceptibility testing observed by Henriksen et al. [3] may result not from true variations in the susceptibility pattern, but rather from the subculture and expansion of different subpopulations after bacterial retrieval from stored samples. This could be a consequence of selecting colonies that differ from the original used in antibiotic testing, or might reflect aspects related to bacterial physiology, such as the limited regrowth capacity of strains with deletions of specific genotypic traits under critical culture conditions, such as those present after cryopreservation. Combined phenotypic and genotypic studies on H. pylori before and after preservation may help in the further definition of the reasons for the discrepancies observed when repeating antibiotic susceptibility tests. Second, one of the primary aims of the clinical microbiologist is to reinforce the positive and negative predictive values of microbiological tests and to help the clinician to optimise antibiotic therapy. The data generated by new clinical and experimental approaches may initially be frustrating, by showing an increasingly complex relationship between microrganisms and man. However, such data may subsequently help to explain discrepancies between laboratory data and in-vivo results. Our own ongoing studies demonstrate that each patient may harbour variable percentages of metronidazole-resistant bacteria among the entire bacterial population. The next step should include an analysis of the possible relationship between the proportion of resistant bacteria and treatment outcome. These data will help us to understand whether treatment failure depends on the ‘load’ of resistant bacteria or on other unknown factors. Molecular biology studies have suggested interesting functional relationships between different genomic traits of H. pylori. In particular, it has been shown that the composition of the cag island greatly affects bacterial motility, the ability to survive in different gastric microenvironments, and the capacity to induce production of proinflammatory cytokines [7–9]. Compared to cagA– strains, cagA+ strains induce more potent interleukin-8
2004 Copyright by the European Society of Clinical Microbiology and Infectious Diseases
1106 Clinical Microbiology and Infection, Volume 10 Number 12, December 2004
production and a greater degree of inflammation, and favour the penetration of antibiotics into the gastric mucosa; bacterial genetic traits may therefore be linked not only to gastric histopathology, but also to the success of antibiotic treatment. In conclusion, multiparametric studies are required to better define the heterogeneous nature of H. pylori strains colonising humans. A combined approach, including the correlation between in-vitro antibiotic testing, in-vivo treatment outcome, and bacterial genotypic traits, may improve our knowledge significantly. P. De Paoli*, M. L. Tomasini and G. Basaglia Microbiology, Immunology, Virology, Centro Di Riferimento Oncologico, IRCCS, Aviano, Italy *E-mail: [email protected]
REFERENCES 1. Alarcon T, Domingo D, Lopez-Brea M. Antibiotic resistance problems with Helicobacter pylori. Int J Antimicrob Agents 1999; 12: 19–26. 2. Jorgensen M, Daskalopoulos G, Warburton V, Mitchell H, Hazell S. Multiple strain colonization and metronidazole resistance in H. pylori-infected patients: identification from sequential and multiple biopsy specimens. J Infect Dis 1996; 174: 631–635. 3. Henriksen T, Lerang F, Lia A et al. Laboratory handling of Helicobacter pylori critically influences the results of in-vitro metronidazole resistance determination. Clin Microbiol Infect 2004; 10: 315–321. 4. Van der Wouden E, de Jong A, Thijs J, Kleibeuker J, van Zweet A. Subpopulations of H. pylori are responsible for discrepancies in the outcome of nitroimidazole susceptibility testing. Antimicrob Agents Chemother 1999; 43: 1484– 1486. 5. Tomasini M, Zanussi S, Sozzi M, Tedeschi R, Basaglia G, De Paoli P. Heterogeneity of cag genotypes in Helicobacter pylori isolates from human biopsies. J Clin Microbiol 2003; 41: 976– 980. 6. Basaglia G, Sperandio S, Tomasini M, Stocco S, Giordari F, De Paoli P. Analysis of antimicrobial susceptibility and virulence factors in Helicobacter pylori clinical isolates. J Chemother 2004 (in press). 7. Suerbaum S, Michetti P. Helicobacter pylori infection. N Engl J Med 2002; 347: 1175–1186. 8. Figura N, Trabalzini L, Mini R et al. Inactivation of Helicobacter pylori cagA gene affects motility. Helicobacter 2004; 9: 185–193. 9. Karita M, Blaser MJ. Acid-tolerance response in Helicobacter pylori and differences between cagA+ and cagA– strains. J Infect Dis 1998; 178: 213–219.
10.1111/j.1469-0691.2004.01035.x
Antimicrobial resistance of Acinetobacter spp. in Europe We congratulate Drs Van Looveren and Goossens for their excellent review regarding the antimicrobial resistance of Acinetobacter spp. in Europe [1]. We agree with their statement that colistin should be considered as a therapeutic option in patients with severe infections caused by multidrug-resistant Acinetobacter baumannii strains. However, we believe that results from recent studies suggest that colistin is less toxic than thought previously and reported in the review. In fact, the reported rates of nephrotoxicity (27% and 58% in patients with baseline normal and abnormal renal function, respectively) do not represent development of ‘renal failure’, as stated in the review [1] and the original paper by Levin et al. [2], but rather renal dysfunction, as evidenced by only a moderate increase in serum creatinine levels. Specifically, it is stated in the original paper [2] that the mean (± SD) increase in serum creatinine concentration was 0.9 ± 0.6 mg ⁄ dL in patients with normal baseline creatinine levels, and 1.5 ± 1.4 mg ⁄ dL in patients with abnormal baseline creatinine levels. In addition, nephrotoxicity did not lead to discontinuation of treatment for any of the patients in that study, and was reversible in a considerable proportion of patients. Other investigators, including our own group [3,4], have noted that the use of intravenous colistin is associated with considerably less nephrotoxicity than reported previously. In addition, Garnacho-Montero et al. [5] reported an almost two-fold nephrotoxicity rate among patients treated for A. baumannii ventilator-associated pneumonia with imipenem when compared to colistin-treated patients [5]. Another point of interest is that methods of colistin administration other than the intravenous route may have clinical value. These include aerosolised (nebulised) administration of the antibiotic to patients with severe nosocomial pneumonia. The value of aerosolised colistin in the prevention and treatment of infections caused by Pseudomonas aeruginosa strains has already been proven for patients with cystic fibrosis. In addition, intraventricular administration of colistin
2004 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 10, 1105–1107
10.1111/j.1469-0691.2004.00985.x
island, although the composition of this gene cluster did not affect resistance to clarithromycin or amoxycillin [6]. Based on these observations, it can be hypothesised that the differences in susceptibility testing observed by Henriksen et al. [3] may result not from true variations in the susceptibility pattern, but rather from the subculture and expansion of different subpopulations after bacterial retrieval from stored samples. This could be a consequence of selecting colonies that differ from the original used in antibiotic testing, or might reflect aspects related to bacterial physiology, such as the limited regrowth capacity of strains with deletions of specific genotypic traits under critical culture conditions, such as those present after cryopreservation. Combined phenotypic and genotypic studies on H. pylori before and after preservation may help in the further definition of the reasons for the discrepancies observed when repeating antibiotic susceptibility tests. Second, one of the primary aims of the clinical microbiologist is to reinforce the positive and negative predictive values of microbiological tests and to help the clinician to optimise antibiotic therapy. The data generated by new clinical and experimental approaches may initially be frustrating, by showing an increasingly complex relationship between microrganisms and man. However, such data may subsequently help to explain discrepancies between laboratory data and in-vivo results. Our own ongoing studies demonstrate that each patient may harbour variable percentages of metronidazole-resistant bacteria among the entire bacterial population. The next step should include an analysis of the possible relationship between the proportion of resistant bacteria and treatment outcome. These data will help us to understand whether treatment failure depends on the ‘load’ of resistant bacteria or on other unknown factors. Molecular biology studies have suggested interesting functional relationships between different genomic traits of H. pylori. In particular, it has been shown that the composition of the cag island greatly affects bacterial motility, the ability to survive in different gastric microenvironments, and the capacity to induce production of proinflammatory cytokines [7–9]. Compared to cagA– strains, cagA+ strains induce more potent interleukin-8
2004 Copyright by the European Society of Clinical Microbiology and Infectious Diseases
1106 Clinical Microbiology and Infection, Volume 10 Number 12, December 2004
production and a greater degree of inflammation, and favour the penetration of antibiotics into the gastric mucosa; bacterial genetic traits may therefore be linked not only to gastric histopathology, but also to the success of antibiotic treatment. In conclusion, multiparametric studies are required to better define the heterogeneous nature of H. pylori strains colonising humans. A combined approach, including the correlation between in-vitro antibiotic testing, in-vivo treatment outcome, and bacterial genotypic traits, may improve our knowledge significantly. P. De Paoli*, M. L. Tomasini and G. Basaglia Microbiology, Immunology, Virology, Centro Di Riferimento Oncologico, IRCCS, Aviano, Italy *E-mail: [email protected]
REFERENCES 1. Alarcon T, Domingo D, Lopez-Brea M. Antibiotic resistance problems with Helicobacter pylori. Int J Antimicrob Agents 1999; 12: 19–26. 2. Jorgensen M, Daskalopoulos G, Warburton V, Mitchell H, Hazell S. Multiple strain colonization and metronidazole resistance in H. pylori-infected patients: identification from sequential and multiple biopsy specimens. J Infect Dis 1996; 174: 631–635. 3. Henriksen T, Lerang F, Lia A et al. Laboratory handling of Helicobacter pylori critically influences the results of in-vitro metronidazole resistance determination. Clin Microbiol Infect 2004; 10: 315–321. 4. Van der Wouden E, de Jong A, Thijs J, Kleibeuker J, van Zweet A. Subpopulations of H. pylori are responsible for discrepancies in the outcome of nitroimidazole susceptibility testing. Antimicrob Agents Chemother 1999; 43: 1484– 1486. 5. Tomasini M, Zanussi S, Sozzi M, Tedeschi R, Basaglia G, De Paoli P. Heterogeneity of cag genotypes in Helicobacter pylori isolates from human biopsies. J Clin Microbiol 2003; 41: 976– 980. 6. Basaglia G, Sperandio S, Tomasini M, Stocco S, Giordari F, De Paoli P. Analysis of antimicrobial susceptibility and virulence factors in Helicobacter pylori clinical isolates. J Chemother 2004 (in press). 7. Suerbaum S, Michetti P. Helicobacter pylori infection. N Engl J Med 2002; 347: 1175–1186. 8. Figura N, Trabalzini L, Mini R et al. Inactivation of Helicobacter pylori cagA gene affects motility. Helicobacter 2004; 9: 185–193. 9. Karita M, Blaser MJ. Acid-tolerance response in Helicobacter pylori and differences between cagA+ and cagA– strains. J Infect Dis 1998; 178: 213–219.
10.1111/j.1469-0691.2004.01035.x
Antimicrobial resistance of Acinetobacter spp. in Europe We congratulate Drs Van Looveren and Goossens for their excellent review regarding the antimicrobial resistance of Acinetobacter spp. in Europe [1]. We agree with their statement that colistin should be considered as a therapeutic option in patients with severe infections caused by multidrug-resistant Acinetobacter baumannii strains. However, we believe that results from recent studies suggest that colistin is less toxic than thought previously and reported in the review. In fact, the reported rates of nephrotoxicity (27% and 58% in patients with baseline normal and abnormal renal function, respectively) do not represent development of ‘renal failure’, as stated in the review [1] and the original paper by Levin et al. [2], but rather renal dysfunction, as evidenced by only a moderate increase in serum creatinine levels. Specifically, it is stated in the original paper [2] that the mean (± SD) increase in serum creatinine concentration was 0.9 ± 0.6 mg ⁄ dL in patients with normal baseline creatinine levels, and 1.5 ± 1.4 mg ⁄ dL in patients with abnormal baseline creatinine levels. In addition, nephrotoxicity did not lead to discontinuation of treatment for any of the patients in that study, and was reversible in a considerable proportion of patients. Other investigators, including our own group [3,4], have noted that the use of intravenous colistin is associated with considerably less nephrotoxicity than reported previously. In addition, Garnacho-Montero et al. [5] reported an almost two-fold nephrotoxicity rate among patients treated for A. baumannii ventilator-associated pneumonia with imipenem when compared to colistin-treated patients [5]. Another point of interest is that methods of colistin administration other than the intravenous route may have clinical value. These include aerosolised (nebulised) administration of the antibiotic to patients with severe nosocomial pneumonia. The value of aerosolised colistin in the prevention and treatment of infections caused by Pseudomonas aeruginosa strains has already been proven for patients with cystic fibrosis. In addition, intraventricular administration of colistin
2004 Copyright by the European Society of Clinical Microbiology and Infectious Diseases, CMI, 10, 1105–1107