By now, the news of a colistin-resistant E. coli isolate from a patient in the United States is widespread, with many major news sources covering the story. Most outlets highlighted the ease of future transfer of the plamid-borne mcr-1 gene between bacteria, the role of agricultural antibiotic use in generating resistant strains, and the looming antibiotic crisis once this gene spreads to carbapenem-resistant Enterobacteriaceae. This particular strain was identified as part of a routine screen for all extended-spectrum beta-lactamase (ESBL)-producing clinical isolates, and was found to harbor fifteen different resistance genes – but fortunately remained susceptible to carbapenem-class antibiotics, and so the patient, a 49-year-old woman, was successfully treated.
The patient had presented with a urinary tract infection (UTI), which is a common infection, especially among women. Over 8.1 million patients visit a health care provider to treat a UTI each year in the United States alone, and 20% of young women with a first UTI will have a recurrent infection. How do providers know if these recurrent infections harbor antimicrobial resistance genes? While culture and an antibiogram are still the gold standards, scientists are hard at work to better predict resistant organisms from previous urine cultures to improve antibiotic usage. An analysis of 4409 patients with nearly 20,000 paired positive urine cultures is now published in Antimicrobial Agents and Chemotherapy.
One of the major problems this study addresses is that patients presenting symptoms want immediate relief, and physicians will often prescribe antibiotics empirically: based on similar cases, before receiving clinical lab results. Empirical treatment isn’t always a bad thing: if it works, the patient feels better and complications that can occur from prolonged UTI are averted. However, if a patient receives a drug to which her or his infection is already resistant, not only is the infection not treated, but the drug exposure can also eliminate other resident microbes, opening up a niche for the invading species. Those that aren’t killed may be selected for resistance themselves.
UTIs are particularly a problem in hospitals, where multidrug-resistant strains are prevalent and the use of indwelling catheters makes patients vulnerable to catheter-associated UTIs (CAUTIs). UTIs are the most common nosocomial infection, and these patients often need treatment as quickly as possible. In healthcare settings, doctors will look at previous lab results, when available, to decide on empirical treatment, but how recent must the lab results be to be helpful? Using retrospective patient data, first author Yaakov Dickstein and senior scientist Michal Paul addressed this important question.
The scientific team first looked at resistant cultures from a patient against the resistance profile of that patient’s immediately previous culture. They then looked for any previously resistant culture from that patient (see figure, right). These were both compared against how common that same resistance is within the general population.
Not surprisingly, a previously resistant culture predicted that the current isolate would retain that resistance phenotype. Probability varied with different types of resistance, however: carbapenem resistance was most likely to be seen repeatedly, while ciprofloxacin resistance was predictive, its predictive power waned with time. For this reason, culture results within the previous six months were found to be most informative for cipro-r, ESBL, or CRNF cultures, though CRE cultures older than six months remained associated with current CRE status.
This study correlates nicely with another retrospective study, which looked at the drug treatment itself. In that study, they found that UTI patients treated with the same antibiotic they’d been treated with during previous episodes were more likely to resolve the infection. Together, these studies emphasize the importance of examining a patient’s records before prescribing medication for UTI.
The mcr-1 positive isolate remained susceptible to carbapenemases, so the patient’s infection was successfully treated. But now that colistin resistance has been detected, health care practitioners may soon have to incorporate regular colistin resistance profiling to stay ahead of the mcr-1 gene dissemination.
-- Julie Wolf