H pylori regimens, stratified by clarithromycin sensitivity

There was a recent systematic review and network meta-analysis of treatments of H Pylori infections in countries with high vs low clarithromycin resistance (see hpylori rx network metaanal gut2016 in dropbox, or doi:10.1136/gutjnl-2016-311868). A network meta-analysis is helpful when there are not head-to-head comparisons of all of the therapies, using mathematical assumptions/manipulations to approximate results across studies in an attempt to approximate the likely results if there were such direct comparisons.  There was another recent blog which did a similar analysis of H pylori treatments but was not stratified by clarithromycin sensitivity (see http://gmodestmedblogs.blogspot.com/2015/08/h-pylori-regimens.html  from the BMJ in 2015). details of the current analysis:

--117 trials with 32,852 patients analyzing 17 H pylori eradication regimens were included

--mean age 48, trials ranged from 58 to 1463 patients, with mean of 281

--all had documented H pylori infections by urea breath test, histologic or bacterial culture exams, or stool antigen

--all had documentation of H pylori eradication by intention-to-treat analysis, at least 4 weeks after therapy

--the authors analyzed results of treatment stratified by clarithromycin resistance, ascertained by a literature review of it prevalence by the countries where the RCTs were done, with a cutpoint of 15% resistance rate.

results:

--compared to a 7-day course of clarithromycin-based triple therapy:

    --overall: best was 14 days of sequential therapy, OR= 3.74 (2.37-5.96)

    --in areas of high clarithromycin-resistance: best was 14 days of sequential therapy, OR= 6.53 (3.23-13.63). and this did not diminish in the more recent studies where resistance was likely to increase; second best was bismuth, PPI and 2 of amoxacillin/tetracylcine/metronidazone/clarithromycin for 10 or 14 days, with OR 3.6

    --in areas of low clarithromycin-resistance: best was 10 or more days of hybrid therapy, OR= 2.85 (1.58-5.37); second best was continuous therapy (eg PPI, amoxacillin and clarithromycin) at least 10 days

    --the lowest efficacy overall was triple therapy with metronidazole for 7 days

    --overall, better efficacy with longer treatments, which mostly meant 14 days

    ​--severe adverse events were rare in all regimens, though happened more often in the longer courses of therapy

Commentary:

--H pylori infection is unbelievably common, in >50% of world's population, and is associated with peptic ulcer disease, active gastroduodenal ulcer bleeding, gastric MALT (mucosa-associated lymphoid tissue lymphoma) and >75% of gastric cancers

--H pylori eradication is recommended in patients on long-term NSAIDs, anti-platelet agents/low-dose ASA, and those with unexplained iron-deficiency anemia or ITP, per the Maastrich IV/Florence Consensus Report in 2012 (see Gut. 2012; 61:646) [as mentioned in prior blogs, see URL at the end of this blog, i saw a dramatic case 20 years ago of a person with intractable steroid-resistant ITP, scheduled for surgery, but the Lancet had a small case report showing benefit of H pylori treatment. so I tested and treated him for H pylori and the ITP melted away, off all meds, within weeks. and without surgery

--the standard therapy has been a PPI, amoxacillin and clarithromycin or metronidazole, but the effectiveness has been waning.

--there are some concerns about generalizing the results of the study above. for example all of the 14 day sequential therapy trials were done in Asia. and the inherent and different biases in the different studies, as well as patient characteristics across studies, make confident conclusions difficult (eg some studies did not even record smoking prevalence), and studies had different levels of blinding

--Countries with high clarithro resistance: China, Croatia, France, Greece, Hong Kong, India, Iran, Italy, Japan, Morocco, Spain, Saudi Arabia, Singapore, Turkey

--Countries with low clarithro resistance: Chile, Ecuador, Finland, Kenya, Korea, Kuwait, Mexico, Taiwan, Thailand, UK, US, Kosovo, Palestine, Romania (last 3 with no recorded resistance rate)​

 --Unfortunately, and a bit surprisingly, it was really hard to find out exactly what the therapies were (they were all combined into categories, such a “sequential” or “hybrid” therapies), even after scouring this article and its supplement. My best guess is the following:

    --sequential therapy for 10 days: first 5 days using PPI (eg omeprazole 20mg) and amoxacillin 1gm; followed by  PPI, clarithromycin 500mg and metronidazole 500mg for the remaining 5 days (some studies have substituted tinidazole 500mg for the metronidazole 500mg). all meds given BID 

    --hybrid therapy for 14 days: first 7 days PPI (eg omeprazole 20mg) and amoxacillin 1gm; followed by PPI, amoxacillin 1gm, clarithromycin 500mg and metronidazole 500mg for 7 days (again, some studies substituting tinidazole 500mg for metronidazole 500mg). all meds given BID

    --so, the hybrid therapy seems to be the same as the sequential therapy, with amoxacillin for the whole course instead of just the first half

--The sequential therapy (which includes clarithromycin at the end) works even with clarithromycin resistance, with 89% of clarithromycin-resistant strains responding vs 29% with standard therapy -- the posited mechanism for resistance being that H pylori had developed efflux channels for clarithro that rapidly remove the drug from the bacteria and that amoxacillin changes the structure of the cell membrane to prevent effective clarithromycin efflux and therefore resistance. (see Ann Intern Med 2007; 146:555-563, and http://gmodestmedblogs.blogspot.com/2013/11/h-pylori-therapy-sequential-vs-standard.html  )​

 --From 9/9/15 blog (see http://gmodestmedblogs.blogspot.com/2015/08/h-pylori-regimens.html ): "and, one very important issue in the US is that we do not have data (at least in Boston) on H pylori sensitivities. maybe that makes sense. one of our greatest assets in Boston is the wonderful ethnic diversity here. one of the problems (at least as H pylori is concerned) is that H pylori organisms from different parts of the world are here​ with their different antibiotic sensitivities. so applying averaged sensitivities may not really benefit the individual patient in front of you."

​--I personally have been using only the 10-day sequential therapy, since it was shown years ago to be superior,  though there are some differences in different studies. this therapy is certainly a bit more complicated: I usually have the patient bring back the bottles of pills and either I or a nurse review the regimen in detail. In the very few cases of persistent H pylori infection post-therapy that I have seen (mostly detected by stool antigen testing), I use the bismuth-based regimen: high dose PPI bid, amoxicillin 1gm bid, levofloxacin 500 mg in the evening, and bismuth subcitrate 240mg bid for 14 days. so far, so good.

--But, based on the current study, I will change to the 14-day sequential therapy course. I am certainly concerned that many of my patients are getting too many antibiotics, esp azithromycin (eg, see http://gmodestmedblogs.blogspot.com/search/label/antibiotic%20resistance ) and metronidazole, which are likely to breed resistance in untreated H pylori infections. so, I follow symptoms closely post-therapy, check H pylori antigen when the patient is symptomatic, and retreat with one of the rescue regimens.  if this resistance happens more often, I will switch to the hybrid regimen above.  also, tinidazole may be more effective than metronidazole, and tinidazole is available in the US but mostly with the disincentive of a Prior Approval....... ​[though, the tinidazole may be more effective in the US precisely because we use much much more metronidazole]

also see http://gmodestmedblogs.blogspot.com/search/label/H%20Pylori​ has a series of blogs on H Pylori infection, including some suggesting eradication leads to lower incidence of gastric cancer, increased bleeding if H Pylori positive and taking NSAIDs (and some studies showing that eliminating H Pylori leads to lower risk of NSAID-associated bleeding), and a review of several studies of different medication regimens.