In a randomized controlled trial at nine hospitals in Taiwan, researchers compared genotypic resistance-guided sequential rescue treatment (n=205) with medication-guided treatment (n=205) in patients with persistent Helicobacter pylori infection who had had two or more failed eradication therapies. Treatment in the former group was based on markers of resistance determined from gastric biopsy specimens, while treatment in the latter was based on medication history. At six or more weeks after completion of therapy, genotypic resistance-guided and medication-guided therapies did not significantly differ for eradicating refractory H. pylori infection (78% vs. 72% eradication rate with first-course rescue therapy; 80% vs. 79% eradication rate after one or two courses, respectively).
Liou and colleagues compared a third-line, 14-day sequential treatment based on antibiotic susceptibility with a similar empirical treatment designed by consideration of previously failed therapies in patients with persistent H pylori infection after 2 eradication treatments. Unexpectedly, they found similar cure rates for both the resistance-guided and the medication-guided groups.
This finding is particularly striking because patients in the trial were not blinded, and both adherence and the placebo effect might be expected to be stronger for genotypic resistance–guided treatment than for empirical treatment. In addition, development of multiresistant strains was extremely frequent after 2 treatment failures, and both groups received a sequential treatment, which we know is less effective than concomitant quadruple therapies for multiresistant bacterial strains. This may have also favored the resistance-guided treatment group, in which patients received only antibiotics known to be effective.
The quality of evidence supporting susceptibility-based first-line therapy is very low, mainly because most comparative RCTs have used suboptimal empirical treatments for comparators. Further, there is no evidence to support susceptibility-based treatment in second- or third-line treatment.
It should also be highlighted that, in the trial by Liou and colleagues, cure rates were low in both groups. There is growing evidence that 14-day concomitant quadruple empirical therapies may cure 90% of the infections, even in patients with multiresistant strains. These therapies will leave little room for improvement by the more expensive and less convenient resistance-guided treatments.