Eight antimicrobial agents were tested against each H. pylori isolate using the antibiotic test discs (amoxicillin, levofloxacin clarithromycin and tetracycline) (Oxoid Ltd). Other drugs are metronidazole, ciprofloxacin, cefuroxime and tinidazole. Inoculum from the subculture was made and streaked onto Brucella Agar containing 7% sheep blood. The antibiotic strips were added and then the plates were incubated at 37 °C in a micro-aerobic environment provided by the CampyGen system (Oxoid Ltd) for another 3–4 days. The antibiotics’ susceptibility (S) and resistance (R) were recorded for each of the antibiotics previously listed.

In the histopathology laboratory, the paraffin-embedded tissue blocks were sectioned at 4 µm thickness and stained with the routine H and E stain for morphology, while modified Giemsa and Warthin-Starry stains were used for the identification of H. pylori. Sections found to show spiral bacteria in the mucosal layer or on the surface of the gastric epithelial cells were considered positive for H. pylori as assessed independently by two histopathologists.

A stool sample was collected from each patient on the day of endoscopy to assess for H. pylori faecal antigen using H. pylori Antigen RapiCard Instatest card (Cortez Diagnostics Inc., Calabasas, CA, United States [US]). If found to be positive, amoxicillin 1 g bid and clarithromycin 500 mg bid along with rabeprazole 20 mg od were prescribed to the patients to take for 14 days, and they were instructed to bring a second stool sample 4 weeks after completing the medications (post-treatment) to assess the eradication or otherwise. The faecal antigen tests were performed on fresh stool samples according to the manufacturer’s direction as described below. The required numbers of sample diluent vials were labelled with the sample identity number and the cap removed. The applicator stick, which was attached to the cap of the stool specimen bottle, was used to transfer a small piece of formed stool into the sample bottle containing specimen preparation buffer, whereas a liquid or semi-solid stool sample (100 mL of sample) was added to the vial using a disposable pipette. The cap was replaced and the mixtures homogenised for 15 seconds by manually shaking vigorously. With the cassette removed from the sealed foil pouch, the sample bottle tip was snapped off and three drops of diluted sample was added to the well. The result was read within 10–15 minutes.

Data obtained were analysed using Statistical Package for Social Sciences, version 16.0 (SPSS Inc., 2009, Chicago, IL, US). Quantitative variables were summarised using range, means and standard deviation, while qualitative variables were summarised using ratios, proportions and percentages. A value of p < 0.05 was regarded as statistically significant. Determinants and predictors were explored using univariate and multivariate analyses with logistic regression.

Ethical approval to conduct the research on antibiotic sensitivity pattern in dyspeptic patients in Kano, Nigeria, was obtained from the Ethical Committee of the Aminu Kano Teaching Hospital, Kano, Nigeria (clearance number: AKTH/MAC/SUB/12A/P3/IV/608).

Out of the 320 samples cultured, 109 grew H. pylori, while the remaining 211 were negative, as seen in Table 2. Of the 109 subjects who were H. pylori culture positive, 55 (50.5 %) were males, while 54 (49.5 %) were females, with male:female ratio of 1:1.

As depicted in Table 3, the sensitivity of the 109 H. pylori isolates cultured is as follows.

All 109 (100%) H. pylori isolates were sensitive to clarithromycin and levofloxacin, but only 10 (9.2%) isolates were sensitive to amoxicillin. None of the isolates were sensitive to metronidazole, tetracycline, tinidazole, ciprofloxacin and cefuroxime.

This study revealed an antimicrobial susceptibility rate of only 9.2% for amoxicillin, an antibiotic that is commonly used in triple therapy for H. pylori eradication in Nigeria (Table 3). Jaka et al. in a systematic review and meta-analysis involving 26 articles reported a 72.6% H. pylori resistance to amoxicillin,¹⁵ and concluded that prevalence of metronidazole, clarithromycin and amoxicillin resistance is high in the developing world including Africa, which, they said, could impair the first-line triple therapy of H. pylori infection. An earlier study in Western Nigeria⁵ reported 34% amoxicillin sensitivity, while in 2007 Aboderin et al. reported 100% amoxicillin resistance in their H. pylori isolates.⁵ Ndip et al. in Cameroon reported 14.4%⁶ and Wu et al. in a study in China¹⁶ reported a rate of 28.1%. However, Ani et al. in North-Central Nigeria,⁴ almost two decades earlier, had reported 100% susceptibility to amoxicillin. This study coming from same part of the country highlighted the likely acquisition of resistance overtime. Other studies from Malaysia, Saudi Arabia and Lebanon reported 100%, 99% and 100% sensitivity to amoxicillin, respectively.^17,18,19 These differences in antibiotic susceptibility in the different countries could be attributed to the differences in local antibiotic prescription practices and usage in the various countries at the time of the studies. In Nigeria, drug regulations are lacking and the few laws enacted are not enforced, thus antibiotics and other drugs are sold over the counter leading to their abuse, thus encouraging the development of resistance in the individuals later. In addition, there is widespread proliferation of cheap and substandard drugs including antibiotics in the country which encourages the development of resistance not only to H. pylori but to other infections as well. Studies have shown that in Africa amoxicillin and ampicillin are the most abused antibiotics in both rural and urban areas because they are cheaply available in oral formulations.²⁰

We recorded 100% sensitivity to clarithromycin (Figure 2) in this study, which was similar to that of Kimanga et al.¹¹ in Kenya, Rizwan et al. in Saudi¹⁸ and Sharara et al.¹⁹ in Beirut, who reported 100%, 96% and 96% clarithromycin susceptibility, respectively. Clarithromycin resistance the world over was reported as low, ranging from 0% to 45%.^18,21,22 Studies from Malaysia¹⁷ reported 97.9% and in the United States^23,24 93.9% sensitivity. However, Kumala et al.²⁵ in Indonesia and another study from Cameroon⁶ reported lower rates of clarithromycin susceptibility of 72.2% and 55.3%. A major contrast to the findings in our study and the others noted above is a study in Western Nigeria,⁵ which reported that 100% of their H. pylori isolates were resistant to clarithromycin. We are not aware of any other study globally that has reported 100% clarithromycin resistance; therefore, the reason for this marked difference is not clear because clarithromycin resistance is very low and as such it is used widely in the treatment of H. pylori infections in most parts of the world.

All 109 (100%) isolates in our study were susceptible to levofloxacin (Figure 2). To our knowledge, there is no study that has tested levofloxacin sensitivity in Nigeria. Norazah et al.¹⁷ in Malaysia reported 99% sensitivity to levofloxacin, while a systemic review²⁶ showed a susceptibility rate of 88.4% in Asia, 75.9% in Europe and 0% in Africa. Levofloxacin can therefore be a suitable substitute to amoxicillin in the triple therapy used for H. pylori eradication in Nigeria.

We found 100% resistance of our isolates to ciprofloxacin as seen in Figure 3; however, a study from Ile-Ife⁵ in Western Nigeria, a decade earlier, showed 15.6% ciprofloxacin resistance, and the resistance rate in Jakarta was 6.9%.²⁵ The reason for this difference is not apparent; however, different antibiotic usage and abuse in the two regions as well as the time of the studies could be responsible for this difference. In Nigeria, there is tremendous pressure on the use of ciprofloxacin by general practitioners for the treatment of typhoid fever, the diagnosis that is often erroneous. Moreover, ciprofloxacin is cheap and readily available in both rural and urban areas, and thus abused by individuals and general practitioners. On the contrary, levofloxacin is new and expensive and as such not subjected to abuse, which may explain the noted susceptibility seen in our study. To our knowledge, there is no other H. pylori study on levofloxacin in Nigeria.

The high resistance rate of 100% observed for metronidazole and tinidazole (Figure 3) is comparable to numerous studies from many developing countries that have reported similar results to our findings. Two separate studies done in Western Nigeria reported 95% and 100% resistance to metronidazole.^5,11 Studies in Cameroon, Indonesia and China reported 93.2%, 63.9% and 100% resistance to metronidazole, respectively.^6,25,27 These findings could be attributed to the poor drug regulation in Africa and other developing countries leading to high antibiotic abuse that encourages development of resistance. Metronidazole resistance has been reported in 10% – 50% of patients in developed countries,²⁸ whereas in developing countries virtually all strains of H. pylori have been found to be resistant to metronidazole.²⁹ The high prevalence of metronidazole resistance in developing countries like Nigeria may also be related to the frequent use of imidazole derivatives for the treatment of gastrointestinal and gynaecological infections often at substandard doses for the relief of diarrhoeal disease which are of frequent occurrence. Osato et al.³⁰ reported that metronidazole resistance was seen more in females (63%) than their male (35%) counterparts. From other studies, metronidazole resistance of H. pylori has been shown to be because of the mutational inactivation of the nitroreductase gene (rdxA gene), which encodes nitro-reductase, an enzyme gene, which encodes nitro-reductase, an enzyme that reduces metronidazole to an active form.²⁹

There was 100% H. pylori resistance to tetracycline in this study as seen in Figure 3, which was similar to two other studies by Oladiipo et al., and Smith et al. in Western Nigeria^5,31 which all reported 100% tetracycline resistance. Tetracycline is not currently used for H. pylori eradication in Nigeria because of its high resistance profile. However, in China and Malaysia^16,17 studies reported resistance rates of 1.2% and 0% to tetracycline, and Sharara in Beirut¹⁹ also reported 2% tetracycline resistance.

In this study, 100% of H. pylori isolates were resistant to cefuroxime (Figure 3); in contrast, a study in China²⁷ reported resistance of only 1.1%. Differences in the antibiotic usage and abuse in the two continents may be responsible for this contrast.

One hundred and one subjects (34.0%) were H. pylori faecal antigen test positive pre-treatment, as shown in Table 5 and Figure 4. Of the 101 initially positive, 9 patients defaulted, while 92 came back for a post-treatment eradication test (rapid H. pylori stool antigen test [Hp SAT]). Of these, 12 (13.0%) were still positive post-treatment (see Figure 5 and Table 6). Smith et al.³¹ reported that 53% of their subjects were still positive post-treatment. The prevalence of H. pylori using the Hp SAT in this study was 31.6%, which is similar to the 36.7% reported by Stella et al. in Lagos-Nigeria.³² Currently available eradication regimens for H. pylori are triple drug combination regimens comprising a PPI and two antibiotic drugs, and eradication rates of 70% – 90% are obtained using these regimens.²¹ The eradication rate of 86.9% achieved by our study agrees with the above finding. The eradication rates of H. pylori have been lower than 80% as a result of increased resistance to H. pylori.³³ Gisbert et al. in a systematic review of 89 studies evaluated the H. pylori stool antigen test and reported a sensitivity of 96% with a specificity of 97%.³⁴ The eradication rate may have been lower if re-endoscopy with H. pylori culture and histology was done. We probably achieved this high eradication level despite over 90% amoxicillin resistance because of the fact that proton pumps inhibitors (PPI) in themselves inhibit H. pylori. This property of PPI and the 100% susceptibility to clarithromycin of our isolates would possibly explain this high eradication rate we recorded in spite of the amoxicillin resistance. It could also be explained that the free drugs (PPI, amoxicillin and clarithromycin) provided by the researcher for the required duration of treatment have encouraged patients’ compliance and thus led to the achieved high eradication rate noted. The apparent high eradication rate mainly from clarithromycin effect without the support of a second antibiotic should not be taken for granted – as guidelines generally recommend – for this may expose this robust antibiotic (clarithromycin) in the future to the development of resistance by the H. pylori bacteria.

The high cost of endoscopy and the discomfort of the procedure prevented patients from agreeing to return for repeat endoscopy and biopsy (only 10 out of 320 patients agreed to return for post-treatment repeat endoscopy). We thus had to resort to in vitro testing using the H. pylori faecal antigen test which correlated well with the culture result (see Table 7) in place of the initially intended in vivo testing of the antibiotic sensitivity using a culture to determine the eradication rate.