Epidemiology of the antibiotic resistance of Helicobacter pylori in Canada

Can J Gastroenterol Vol 14 No 10 November 2000 879 Division of Gastroenterology, McGill University Health Center, Montreal, Quebec Correspondence and reprints: Carlo A Fallone, Division of Gastroenterology, Room R2.28, Royal Victoria Hospital, 687 Pine Avenue West, Montreal, Quebec H3A 1A1. Telephone 514-843-1616, fax 514-843-1421, e-mail cfallone@lan1.molonc.mcgill.ca Received for publication May 9, 2000. Accepted June 13, 2000 REVIEW


Épidémiologie de l'antibiorésistance d'Helicobacter pylori au Canada
CONTEXTE : Le taux de résistance d'Helicobacter pylori aux antibiotiques détermine le taux de guérison des schémas thérapeutiques se composant de ces antibiotiques. BUT : Passer la documentation en revue pour déterminer le taux de résistance d'H. pylori au métrodinazole et à la clarithromycine au Canada et vérifier si le taux varie selon les régions. MÉTHODE : On a procédé à un examen exhaustif de la documentation pour déterminer la prévalence de la résistance d'H. pylori au métrodinazole et à la clarithromycine au Canada en cherchant dans MEDLINE de 1980 à mai 1999 et en lisant les résumés de l'American Gastroenterology Association Digestive Disease Week, de la Semaine canadienne des maladies digestives et des European H. pylori Study group Meetings de 1995 à mai 1999. RÉSULTATS : On a trouvé 11 études dans lesquelles il y avait une évaluation du taux de résistance d'H. pylori au métrodinazole et neuf, à la clarithromycine au Canada. Les taux de résistance au métrodinazole et à la clarithromycine variaient de  H elicobacter pylori plays a significant role in the pathogenesis of peptic ulcer disease, gastric adenocarcinoma and mucosa-associated lymphoid tissue lymphoma (1)(2)(3). A suggested approach to the treatment of adult patients younger than 45 years old with uninvestigated dyspepsia and without alarming symptoms is to test for H pylori infection and treat the patient if this infection is present (3). Hence, treatment of H pylori, which has been estimated to be present in approximately 20% to 50% of the Canadian population (4), is increasingly recommended.
The successful cure of H pylori infection is dependent on the susceptibility profile of the organism. It is clear that, although certain strains that show in vitro resistance to metronidazole or clarithromycin respond to treatments containing these antibiotics, the success rate of eradication is significantly reduced (5,6). It is thus important to know the susceptibility profiles in each community to recommend the optimal treatment, given that these in vitro resistance rates vary tremendously from one part of the world to another (7).
The present article summarizes the many studies that have documented in vitro susceptibility profiles of H pylori strains across Canada and attempts to estimate the true rate of primary (ie, before any treatment directed toward H pylori) H pylori resistance to metronidazole and clarithromycin.

METHODS
A MEDLINE search was performed for articles published from January 1980 to May 1999. Abstracts from the Canadian Digestive Disease Week, the American Gastroenterology Association Digestive Disease week and the European H pylori Study Group meetings from 1995 to 1999 were also searched. In addition, authors across Canada who were known to have worked in the field were asked for any published or upcoming data. The search was for any documentation of H pylori susceptibility profiles to any antibiotics. The articles or abstracts identified were then retrieved and analyzed to determine the population studied, whether the study comprised an adult or pediatric population, the diagnosis of the population studied, whether the study comprised previously untreated patients or patients who had previously undergone treatment for H pylori eradication, the method used to determine susceptibility, the cutoff used to define a resistant strain and the estimate of resistance. Using the data obtained from these papers, 95% confidence intervals were calculated throughout. If there were any data missing or if certain data were unclear, an attempt was made to contact the author for clarification.

RESULTS
Eleven studies that estimated H pylori resistance to metronidazole in Canada were identified. Nine studies estimated resistance to clarithromycin in Canada, and five studies examined resistance to other antibiotics. Only one study identified antibiotic susceptibility profiles in strains obtained from a Canadian pediatric population. Some of the studies examined susceptibility profiles in specific regions or cities in Canada, whereas other studies obtained their strains from different cities in Canada. Results are presented as per region, in the event that there was any difference in the rates from different regions of Canada. Resistance to metronidazole: Four studies examined the prevalence of H pylori resistance to metronidazole in the Montreal region (8)(9)(10)(11). Of these studies, two (8,9) were of duodenal ulcer patients specifically. The four studies were of adult populations, and strains were obtained before the patients had undergone any treatment for H pylori. Agar dilution, with a cutoff of 8 mg/L, was used in three studies (8,9,11), whereas the E-test (cutoff 8 mg/L) and disk diffusion (cutoff 20 mm) were used in the other study (10). Resistance to metronidazole was estimated to be 17% (n=119) (9), 20% (n=89) (8) and 11% (n=55) (11) in the studies that used agar dilution, and a higher estimate of 29% to 34% was obtained in the study that used the E-test and disk diffusion (10). The results are graphically demonstrated with their 95% confidence intervals in Figure 1.
One study examined H pylori resistance to metronidazole in Alberta by using both the agar dilution and the E-tests (12). The cutoff was 8 mg/L, and the strains were obtained before any treatment attempt had been made. Strains were from adult patients who had H pylori gastritis. The prevalence of resistance to metronidazole was found to be 12% when agar dilution was used and 14% when the E-test was used (n=42).
Six studies (13)(14)(15)(16)(17)(18) Figure 1). One abstract (18) that did not specify which method was used obtained a resistance rate of 22% (n=100), which is in keeping with results from the majority of the studies (Figure 1). The other two studies (16,17) used the E-test and obtained rates that were higher than other estimates -48% (n=64) and 38% (n=77), respectively. Neither study specified whether patients had previously undergone treatment for H pylori eradication. The studies labelled in Figure 1 as not necessarily excluding patients who had undergone prior treatment for H pylori were performed across Canada. These studies are often multicentre eradication trials, and most eradication trials allow patients who have had one previous attempt at H pylori treatment. Hence, the estimate of H pylori primary resistance would be falsely raised in that study group. This is demonstrated in Figure 1, where the studies that did not allow patients who had undergone treatment attempts seem to have lower estimates of resistance than those that may have included previously treated patients. As well, studies that used E-tests tended to have higher estimates than those that used agar dilution. Therefore, the rate of resistance to metronidazole can be estimated from Figure 1. If the above explanations for discrepancies are accepted, the prevalence of resistance to metronidazole is in the range of 18% to 22% across Canada. Resistance to clarithromycin: One study examined H pylori resistance to clarithromycin in the Maritimes. This study used the E-test on strains obtained from adults with peptic ulcer disease or with nonulcer dyspepsia. These patients had not undergone a previous treatment attempt, and the rate of H pylori resistance to clarithromycin was 1.8% (n=162) (19). In the Montreal region, the same four studies that identified resistance to metronidazole estimated resistance to clarithromycin. The prevalence of resistance varied from 0% to 2% (8)(9)(10)(11). In Alberta, Taylor et al (12) estimated that resistance to clarithromycin was present in 3% (n=31) of patients with H pylori gastritis who previously had not undergone H pylori eradication treatment (12). Three studies obtained strains from across Canada (14,15,17). Each study obtained strains from Montreal, Edmonton and Toronto. Two studies estimated H pylori resistance to clarithromycin to be 2.1% (n=97) and 1.3% (n=77), respectively (15,17). Fedorak et al (14) estimated resistance to clarithromycin using microbroth dilution and found a very high rate of resistance (11.6%, n=147). It is not clear whether some of the patients had undergone a previous attempt at H pylori eradication, potentially accounting for this elevated rate. Figure 2 demonstrates the 95% confidence intervals around the point estimates for each of these studies. If the one outlying estimate is excluded because it likely does not truly represent primary resistance (14), the rate of resistance to clarithromycin across Canada is roughly estimated to be less than 4%. Other antibiotics: Several studies (8)(9)(10)(11)(12)14) have looked at H pylori susceptibility profiles to other antibiotics. No resistance was found to amoxicillin or ampicillin, except for one study that found that three of 147 strains were resistant to amoxicillin (2%) (14). One of the three studies that examined azithromycin and roxithromycin resistance found a 2% resistance rate, whereas the other studies found no resistant strains. No resistance was found in one study that searched for erythromycin resistance nor in the several studies that examined tetracycline resistance.
Pediatric population: Only one study examined the resistance profile of H pylori obtained from a pediatric population in Canada (20). The study was of patients with chronic gastritis who had not previously undergone treatment for H pylori eradication. Agar dilution was used, and all of the 18 strains were susceptible to metronidazole, tinidazole, erythromycin, azithromycin, ampicillin, doxicycline and ciprofloxacin. One of the 18 strains was resistant to cefixime.

DISCUSSION
This article attempts to summarize the several studies that have estimated H pylori resistance to metronidazole, clarithromycin and other antibiotics in Canada. The studies have been analyzed in such a way as to determine the cause of any discrepancies among different estimates. It appears that results obtained using the E-test may be slightly overestimated compared with results obtained using agar dilution. For this reason it is important that future studies adhere to recommendations of the National Committee for Clinical Laboratory Standards (21). In addition, all the studies that tended to have higher estimates of resistance with either metronidazole or clarithromycin were from studies in which the exact population studied was not specified. Namely, it was not specified whether strains were obtained from patients who had previously undergone a failed attempt at H pylori treatment. When treatment with a regimen containing metronidazole fails, a higher number of metronidazole-resistant strains remain (22). The same is true for clarithromycin (23). Hence, the slightly higher estimates of resistance may be due to the inclusion of patients who may have secondary rather than primary resistance.
The region of Canada from which strains were obtained did not seem to affect the rate of resistance. However, most patients in these studies were urban dwellers. We cannot comment on rates of antibiotic resistance in rural dwellers or in certain populations of Canada such as the Native Indians and the Inuit. Studies of strains obtained for these populations would be interesting.
Therefore, in Canada, the prevalence of primary resistance to H pylori is estimated to be approximately 20% for metronidazole and approximately 1% to 4% for clarithromycin. The rate of resistance to other macrolides such as azithromycin and roxithromycin appears to be similar to that of clarithromycin. Resistance to other antibiotics is extremely rare. Resistant H pylori strains in the pediatric population also appear to be rare.