While the news is abuzz with reports of Zika virus coming to a place near you, an interesting article on a much more mundane topic – the treatment of uncomplicated UTIs – should be competing for your attention. It is a publication we have been waiting for, as mentioned in an earlier blog. Results from a small pilot study comparing antibiotic therapy to analgesic therapy (ibuprofen) in acute cystitis patients were intriguing: similar relief of symptoms was seen in both arms, with the antibiotic-sparing regimen demonstrating no harm from withholding microbe-directed therapy [1]. A more definitive study was in the works, and the publication is out now.
In a remarkable first, Gagyor et al. compare causative with symptomatic therapy for uUTI
Now the BMJ gives us detailed results from this more definitive trial [2]. In this much larger well-controlled study of 494 women presenting with acute cystitis in an outpatient setting, antibiotics were not needed by 67% of women. However, without antibiotics it took longer to get over an episode of uUTI. Patients assigned to ibuprofen also were more symptomatic with dysuria and frequency/urgency than those started on antibiotic right away. In other words, the ‘burden of symptoms’ as the authors call it was increased.
The study was well-designed, multi-center, placebo-controlled, double-blind, and adequately powered. Patients with clinical symptoms of acute cystitis were eligible for inclusion; treatment consisted of either ibuprofen at a dose of 400 mg tid x 3 days or a single dose of fosfomycin (3 gram). Patients presenting with signs/symptoms of pyelonephrits were excluded. Follow-up was for 28 days and symptoms were recorded by patients using a reasonable (but not yet validated) scoring system.
This trial has all the trappings of a landmark study
As so often, depending on point of view, proponents of standard antibiotic therapy will find their preference confirmed, and proponents of ‘watchful waiting’ will now have data justifying their approach to uUTI treatment.
The designers of the study took a practical approach and enrolled patients based on symptoms without waiting for culture results. Thus, the primary analysis is based on the ITT population. These are reasonable design features; after all, this was not a registration study which had to follow guidelines (which no longer exist for this indication). The authors chose to use standard 80-125% bioequivalence margins as a test of non-inferiority. This seems a reasonable choice as well.
However, there are a few questions about the inclusion criteria and how they influenced outcomes. Only approx. 76% of women had positive urine cultures at entry. By ITT analysis, only 33% of women in the ibuprofen arm were in need of antibiotic therapy but this percentage went up for the subset of patients with positive urine cultures (defined as >102 CFU/mL) who needed antibiotic therapy in 49%. We would have liked to have seen further breakdown of the data based on the older cut-off of ≥105 CFU/mL, predicting that this would be identify a population with an even higher percentage of antibiotic need. Similarly, patients with a higher symptom score at presentation may have done better with antibiotics than with ibuprofen. Hence, stratification by initial symptom score may be way to discern which patients should be started on antibiotics right away.
This was not a placebo-controlled study. A small but not negligible percentage of patients on ibuprofen required hospital admission for serious AEs unrelated to infection, including a patient with significant GI bleeding. There were also a few patients that progressed to pyelonephritis in the ibuprofen but not in the fosfomycin group.
Comparing symptom improvement scores is a tricky business. Should symptom relief with an analgesic or spasmolytics not be offered equally to patients in both treatment arms? Was concomitant analgesic use captured in the fosfomycin arm? We could not find answers to these questions in the publication nor in the published protocol [3]. Hopefully, the authors will be able to address these and other questions in a subsequent publication.
This study had a long gestation period; it was planned and executed with great attention to detail. The authors offer data sharing upon request, a very generous offer which should become the rule. Funding was provided by a government research fund. Well done – Congratulations!
PLEASE TAKE A MOMENT FOR THIS TOTALLY NON-SCIENTIFIC POLL
[poll id=”6″]
References:
[1] Bleidorn J. Symptomatic treatment (ibuprofen) or antibiotics (ciprofloxacin) for uncomplicated urinary tract infection? – Results of a randomized controlled pilot trial BMC Medicine 2010, 8:30
[2] Gagyor I. Ibuprofen versus fosfomycin for uncomplicated urinary tract infection in women: randomised controlled trial. BMJ 2015;351:h6544
[3] Gagyor I. Immediate versus conditional treatment of uncomplicated urinary tract infection – a randomized-controlled comparative effectiveness study in general practices. BMC Infectious Diseases 2012, 12:146
