The key determinants of the outcome of eradication therapy for Helicobacter pylori infection are compliance and the presence of pretreatment antibiotic resistance of the isolate. The paper by Chang et al. in this issue of the journal draws our attention to the ongoing need to define resistance rates locally and to monitor changes in antibiotic resistance rates as a guide to treatment.1
Limitations of the antibiotic resistance data
Unfortunately, there are major problems with the worldwide and regional data that inform us about antibiotic resistance by H. pylori. Unlike other important infectious diseases, where there are national and international bodies that closely and carefully survey trends in the prevalence of antibiotic resistance, monitoring of H. pylori antibiotic resistance is mostly done in an ad hoc manner. It is usually left to motivated and interested individuals in single tertiary centers to report point prevalence and time trends in resistance rates. Although these reports are helpful, by necessity, such samples are often relatively small, use varying laboratory methods and involve tertiary referral patients that may not be representative of countries with large and heterogeneous populations nor relevant to nearby countries where no data are available and patterns of antibiotic use differ. Where contemporaneous data are available from a number of centers within a single country, any regional variability can be assessed and a better estimation of the scale and patterns of antibiotic resistance determined.2,3
There is a growing opinion in the published work that primary clarithromycin resistance (CR) is rising in some countries.4 However, the evidence for this is confounded by all the methodological issues described above. In addition, there is potential for confusing primary with secondary CR in tertiary centers, as fewer treatment-naive patients and more treatment failures are referred to major centers in recent years. While it is plausible (and perhaps likely) that resistance rates have risen or may be rising in some places, systematic, ongoing monitoring is required to determine this and thus to inform choices for treatment.
Community exposure to antibiotics and primary resistance
There is evidence that primary resistance by H. pylori for some antibiotics is related in part to the community exposure to these antibiotics as monotherapy for other indications. Where clarithromycin has been long used as monotherapy, usually for respiratory infection, higher H. pylori primary CR rates have been reported.3 In France, for example, which is often reported as having the highest primary CR rates in Western Europe, clarithromycin has been available as monotherapy for respiratory and other infections for many years and has been widely used. Similarly, in Australia, somewhat surprisingly, high primary metronidazole resistance (MR) rates have been reported consistently from single centers and from multicenter surveys.2,5,6 It transpires that there is a very large community exposure to monotherapy with nitroimidazole agents (predominantly metronidazole and tinidazole); nearly 1 million prescriptions for such compounds are written annually in a country of just 21 million people.7 Many randomized controlled studies in Australia and elsewhere have documented the adverse impact on outcomes of primary MR on metronidazole containing triple therapies against H. pylori. Given the known high MR rates, these combinations are no longer recommended for first-line therapy in Australia.2–6
In many Asian countries and regions, the community exposure to antibiotics may be harder to ascertain, but given the availability and ease of access to antibiotics at the community level in many countries, it may be considerable also. Certainly, significant resistance has been reported from hospital surveys in many countries in the region.8–10 In the paper by Chang et al., evidence is presented for a rising rate of primary levofloxacin resistance (LR) in Taiwan in recent years, perhaps related to an increasing use of quinolones in that location. As elsewhere, the evidence for a temporal relationship is circumstantial (but plausible) and in this study the timespan was very short, so more longitudinal data are needed. However, it emphasizes the clear need for national surveillance strategies to monitor H. pylori antibiotic resistance rates to inform local treatment guidelines, given the differences in resistance rates between countries.
As primary CR remains relatively low in most countries, recommended first-line treatment for H. pylori remains clarithromycin-based triple therapy, usually combining this agent with amoxicillin and a proton pump inhibitor (PPI).4,11 This recommendation is based on the evidence from a large body of outcome data that includes multiple, multicenter, large randomized controlled trials (RCT) and meta-analyses; the latter provide the highest level of evidence. This evidence base should be contrasted with data that underpins recent commentary and opinion on the need to move away from clarithromycin-based triple therapy for first-line treatment. More recent outcome data have been mostly retrospective single centre, non-comparative audits of treatment outcome in clinical practice. These ‘real world’ reports are and always have been important, but cannot be used to compare outcomes with past RCTs, as results are usually inferior with such methodology. When primary CR rates are known to be high (>15–20%), avoiding clarithromycin first line is reasonable, but in other circumstances such a call seems premature.11
The support for the use of sequential therapy is an example of the clamor for alternative first-line therapies, following some favorable reports of the outcome of this treatment. However, as highlighted in a recent systematic review on this treatment, there is clear evidence of publication bias, as most of the supportive data comes from very few investigators and few geographic locations.12 Furthermore, the combination has not been tested in rigorous randomized comparative trials (as was the case with current established treatments), nor evaluated by a broad range of investigators. While sequential therapy may find a role as an alternative first-line therapy, there are insufficient data to recommend it at present, This view is reflected by the recent Asia–Pacific Consensus guideline on H. pylori to be published soon in the Journal.4
Place of levofloxacin in first- and second-line therapy
Levofloxacin, which is emerging as a useful alternative antibiotic in combination with amoxicillin and a PPI, needs to be considered in the same light when considered as part of an alternative first-line therapy. Randomized controlled trials are needed, but given the commercial realities of H. pylori therapy, these may not occur. It may be that recommendations on the place of levofloxacin triple therapy (and sequential therapy) will ultimately rely on lesser evidence than what has been available previously. Encouragingly, primary levofloxacin resistance appears fairly low in many places where surveys have been done.1,13 However, as with clarithromycin, secondary resistance emerges readily after failed first-line therapy. Outcome data from audits of salvage eradication therapy have been generally favorable.14 However, there is little direct comparison with clarithromycin-based triple therapies to determine whether or not it should supplant clarithromycin as first-line treatment. Individualized decisions to use levofloxacin preferentially may be made for patients who are known to have had prior exposure to clarithromycin as monotherapy, or in that minority who are found to have primary CR after culture of an endoscopic biopsy. As culture is neither available nor practical in all patients, this will only influence treatment choice in a minority of patients. For second-line and subsequent therapy, there are now many single center reports and a meta-analysis of the outcome of levofloxacin triple therapy (usually combined with amoxycillin and a PPI). This is sufficient evidence to accept that levofloxacin has a place as a salvage therapy.14
Thus, first-line therapy for H. pylori infection at present should remain standard PPI, amoxicillin and clarithromycin in communities where the CR rate is known to be modest.4,11 In the penicillin allergic patient, a PPI, nitroimidazole and clarithromycin regimen is often advised and has reasonable eradication rates, although outcomes are adversely affected in regions where primary MR is high and dual secondary MR and CR is common after treatment failure. A PPI, bismuth, tetracycline and metronidazole quadruple therapy is an alternative choice for both first- and second-line therapy for penicillin allergic (and all other) patients but has problems related to access and tolerability. Its main advantage is that the addition of a PPI to standard bismuth triple therapy overcomes pre-existing metronidazole resistance.2 Another option is to formally assess penicillin sensitivity in patients with an uncertain history. In a recent report, more than 80% of people with a history of penicillin allergy were found not to be penicillin allergic after formal testing. They were then able to have amoxicillin containing regimens, thus allowing the use of the best first-line therapy and a greater choice of second-line therapy when needed.15
After failure of first-line clarithromycin triple therapy, there are a number of choices for second-line or subsequent therapy. There are supportive data for these but no large scale randomized comparative data. The best available evidence supports at least three salvage therapies, including PPI, amoxicillin, levofloxacin triple therapy,14 a PPI, amoxicillin, rifabutin triple therapy,16 or quadruple therapy,4,11 in addition to a clarithromycin-based therapy if that antibiotic was not used initially. In patients who have access to culture and sensitivity testing and are found to harbor a metronidazole-sensitive organism, a PPI, amoxicillin, metronidazole combination is also an option. There are therefore several choices of salvage eradication regimens that can be used sequentially if need be. The order in which they should be used after first-line failure is not clear. What is abundantly clear, however, is that poor results are obtained by repeating the same initial treatment. This is particularly the case with clarithromycin re-treatment, for which eradication rates of less than 10% are common (as secondary CR is usually 70–80%). Unfortunately, it is still common practice in some areas for the same clarithromycin-based first-line therapy to be prescribed again, either due to a lack of knowledge or to the absence of a readily available alternative.
In summary, there need to be more resources devoted to systematically monitoring trends in H. pylori antibiotic resistance. These data will inform local treatment recommendations. It remains important to use the most effective treatment first, so as to produce fewer treatment failures and less secondary resistance. Until there is evidence to the contrary, clarithromycin-based triple therapies remain the treatment of choice. There are several choices for second-line of therapy, among which levofloxacin triple therapy is certainly one that is promising. It may find a role in first-line therapy particularly if primary CR rises. For all H. pylori therapies, the quality of the supportive evidence and a sound knowledge of antibiotic resistance trends are critical to decision making. Until a treatment is of proven benefit, we should be mindful of the words of William Osler, who said, ‘I always use the newest medicine first, before its effectiveness wears off’.