Study tests five-antibiotic combos against E. coli
- UCLA researchers reported in a September 3, 2018 study that four- and five-antibiotic combinations killed resistant E. coli more effectively than expected. - The team tested 18,278 combinations from eight antibiotics; 6,443 five-drug combinations and 1,676 four-drug combinations outperformed expectations, UCLA said. - The findings appeared in npj Systems Biology and Applications; treatment guidelines and routine clinical practice have not changed based on this study.
Researchers at the University of California, Los Angeles reported in 2018 that resistant *E. coli* could sometimes be suppressed more effectively with four- and five-drug antibiotic combinations than with the one- or two-drug regimens more commonly discussed in practice. The work, published on September 3, 2018 in *npj Systems Biology and Applications*, is drawing renewed attention after a May 19, 2026 Contagion Live item highlighted the findings. The study was a lab experiment, not a clinical trial, and it did not announce any change to treatment guidelines. What it did show was that higher-order combinations can produce effects that standard assumptions may miss. ### What exactly did the researchers test? UCLA scientists used eight antibiotics and measured how every possible two-, three-, four- and five-drug combination affected *E. coli* growth, for a total of 18,278 combinations, according to UCLA’s summary of the study. The team compared the observed effect of each mix with what they expected based on the drugs’ individual and lower-order effects. (contagionlive.com) The September 3, 2018 paper focused on drug interactions in a bacteria system rather than patients. That distinction matters because a combination that looks strong in vitro still has to clear questions about dosing, toxicity, pharmacology and feasibility before it can be used in people. ### What were the most important numbers? The UCLA team found that 1,676 four-drug combinations performed better than expected and that 6,443 five-drug combinations did the same, according to the university and Contagion Live. (contagionlive.com) The researchers also found many combinations that performed worse than expected: 2,331 four-drug combinations and 5,199 five-drug combinations underperformed. (nature.com) Pamela Yeh, one of the study’s senior authors and a UCLA professor, said in the university release that there had been “a tradition of using just one drug, maybe two.” Van Savage, the other senior author, said he was surprised by how many effective combinations appeared as the number of drugs increased. Those comments framed the paper as a challenge to the idea that adding more antibiotics would usually yield only marginal benefit or canceling effects. (contagionlive.com) ### Why would more drugs help at all? Contagion Live reported that some combinations appeared to work because different antibiotics hit *E. coli* through different mechanisms, including attacks on the bacterial cell wall and DNA. In combination work, that kind of mechanistic spread can matter because one drug may expose vulnerabilities that make another drug more effective. (newsroom.ucla.edu) The broader antimicrobial-resistance literature also describes combination therapy as a strategy researchers use to improve efficacy and manage resistance, though outcomes can be synergistic, neutral or antagonistic depending on the drugs and organism involved. That is consistent with the UCLA result that many combinations improved performance, while thousands also did worse than expected. ### Does this mean doctors are about to use five antibiotics for *E. coli* infections? (contagionlive.com) The May 19, 2026 Contagion Live report did not say any guideline-writing body had adopted five-antibiotic regimens for resistant *E. coli*. The article presented the work as a research finding about what may be possible, not as a new standard of care. Current treatment decisions for serious *E. coli* infections still depend on the site of infection, susceptibility testing, toxicity tradeoffs and resistance mechanisms. (academic.oup.com) A lab signal is an early step, not a prescribing instruction. ### So what should readers watch next? The next concrete step would be follow-on studies that move beyond petri-dish growth measurements into animal models, dosing work or human clinical testing. (contagionlive.com) The relevant published anchor remains the September 3, 2018 paper in *npj Systems Biology and Applications*, and the May 19, 2026 Contagion Live article is the current report that resurfaced it. (nature.com)
