The effectiveness of traditional Helicobacter pylori therapies has declined over the past 10 years coincident with an increase in antimicrobial resistance. Indeed, clarithromycin-containing triple therapy is now not generally recommended for empirical use unless clarithromycin resistance is rare locally or pretreatment clarithromycin susceptibility has been confirmed (that is, only as a tailored therapy). Tay, Marshall and colleagues have now reported the results of tailored multidrug therapies in patients who have previously failed traditional 7-day triple clarithromycin-containing therapy.1 Susceptibility testing and penicillin allergies were used to select the treatment regimen for each patient. The preferred regimen was a four-drug hybrid sequential therapy for 10 days consisting of the PPI rabeprazole and high-dose amoxicillin for 10 days with the addition of rifabutin and ciprofloxacin during treatment days 6–10 (referred to as PARC therapy; see Box 1 for dosage information). Those patients with a penicillin allergy received the PPI plus bismuth subsalicylate for 10 days and rifabutin with ciprofloxacin during days 6–10 (referred to as PBRC therapy; Box 1). Finally, those patients with pretreatment ciprofloxacin and/or rifabutin resistance received individualized tailored therapy (Box 1) based on antibiotic susceptibility testing: typically a 10-day nonsequential course of a PPI, bismuth subsalicylate, furazolidone and one additional antibiotic (namely, tetracycline or metronidazole).

PARC therapy was successful in 95.2% (95% CI 91–97%) of the 210 patients who received it and PRBC therapy in 94.2% (95% CI 85–98%) of 69 patients.1 Success with individualized tailored therapies for those with rifabutin and/or ciprofloxacin resistance (specifically, PBRF, PBFT, PBAC and PRB; Box 1) was 100%; however, there were only 1–4 patients per group. Therapy with PBAF (n = 15), PMA (n = 3), and PBAT (n = 2) was less than 75% effective. The researchers suggested that the two sequential regimens PARC and PBRC could be confidently given to those with known susceptible infections.1

Here, we address three questions: are their results applicable to current clinical practice? As susceptibility testing is not widely available, can the results be used as empiric regimens? Finally, are any of the regimes ready for 'prime time' or is further development needed?

In our opinion, neither four-drug sequential therapy is ready for prime time...

In our opinion, neither four-drug sequential therapy is ready for prime time, especially when in the same region one would expect to cure nearly 100% with a 14-day course of bismuth, tetracycline, PPI and either metronidazole or furazolidone quadruple therapies without risking increasing fluoroquinolone or rifabutin resistance in the non-H. pylori bacterial populations.2,3

Ideally, the rational use of any regimen requires knowledge regarding its effectiveness with susceptible strains as well as the effects of resistance to each of its antimicrobial components. With these data, one can generally assume that knowledge gained from any population with a similar pattern of resistance will be directly transferable to a similar population in any location. Ideal regimens will have also been optimized for each component in relation to: formulation, doses, dosing intervals and relation to meals, and duration of therapy; the modifications to these components needed for resistant infections will have also been identified (for example, 7 days of bismuth-containing quadruple therapy is adequate for susceptible strains whereas 14 days might be optimal when resistance exceeds 40%4).

The suggested sequential regimens contain both ciprofloxacin and rifabutin. However, it is not clear if both drugs are necessary and, moreover, the data presented suggest that only one of the drugs was sufficient. For example, in the study by Tay et al.1 10-day courses using only either ciprofloxacin or rifabutin (such as, PBAC or PBR) provided 100% cure rates with the caveat that with each regimen only one individual was tested. However, Borody et al.5 had previously reported eradication rates of >90% using a rifabutin triple therapy with high-dose amoxicillin (that is, high-dose PPI, 1 g or 1.5 g amoxicillin three times daily, and 150 mg rifabutin daily for 12 days) among Australian patients, also suggesting that ciprofloxacin is redundant.

Besides confirmation that rifabutin and ciprofloxacin both contributed considerably to the outcome, the results presented in the Tay et al.1 study suggest that the efficacy of other combinations, including PBAC, PBR and PBAR, should be examined further. In addition, after deciding on the preferred regimen or regimens among the various therapies tested, optimization would still be needed, especially in terms of doses and durations (for example, would 14 days be better than 7 or 10 days?) and to determine the effect of resistance on outcome (that is, what proportion of resistance will drop the treatment success below 90%?).

The specific drugs used also requires consideration as, for example, rifabutin use is limited in many countries with the goal of 'saving' it for use with tuberculosis, and fluoroquinolone use has tended to focus on second-generation drugs. The prevalence of fluoroquinolone resistance has also increased rapidly worldwide, which has generally undermined the use of this class of drugs for empiric anti-H. pylori therapy. As with other infectious diseases, one would prefer to base therapeutic decisions on knowledge of local or patient-specific resistance patterns. Generally, neither fluoroquinolone nor rifabutin resistance can be overcome by increasing the dosage or duration of therapy, suggesting that the two new sequential therapies have limited usefulness except as tailored therapies. Whether second-generation quinolones would provide equivalent or better results also needs to be explored.

A number of other new, highly successful and promising regimens have been introduced by other investigators, including a 5-day sequential therapy (comprising high-dose PPI, esomeprazole 40 mg twice daily, amoxicillin 1 g twice daily, levofloxacin 500 mg twice daily, and tinidazole 500 mg twice daily),6 a 14-day sequential-concomitant hybrid therapy,7 and a bismuth-containing 14-day sequential therapy programme consisting of pantoprazole (40 mg twice daily for 14 days), colloidal bismuth subcitrate (600 mg twice daily for 14 days) with amoxicillin (1 g twice daily for the first 7 days) and tetracycline (500 mg four times daily) and metronidazole (500 mg three times daily for days 8–14).8 The results reported by Federico et al.6 of the sequential fluoroquinolone-containing regimen contrast with the majority of studies using a PPI, amoxicillin and a fluoroquinolone for 7 or 10 days as they have rarely achieved eradications of 90% or greater.9,10 Whether a ciprofloxacin-containing sequential triple therapy based on the Tay et al.1 study protocol would also be effective other than as a tailored regimen remains to be proven.