Meanwhile, physicians are still choosing which drugs to use, based on a test that hasn’t really changed since John F. Kennedy was president. Moreover, the half century-old test has a serious flaw that’s long been overlooked by doctors: Bacteria act differently inside humans than they do in the lab, which means lab tests can deliver misleading results.
In the past, this wasn’t a big deal because doctors had a potent array of antibiotics at the ready. But those drugs—considered miraculous after Alexander Fleming discovered penicillin in 1928—have begun to lose their magic. Bacteria are increasingly becoming resistant, in part thanks to overuse, leaving doctors with few weapons to quell these infections, also known as superbugs. Such afflictions sicken at least 2 million Americans a year and kill 23,000, according to a rough estimate from the Centers for Disease Control in 2013. A British government report (PDF) in 2014 projected that, by mid-century, superbugs will kill more people than cancer.
In the face of this growing crisis, scientists are looking for better tests to identify which drugs will work.
“Basically, the technology is 50 or 60 years old," said Henry Chambers, an infectious disease specialist and professor at the University of California-San Francisco School of Medicine. “Maybe what we thought was the gold standard [test] may not be the gold standard after all."
“We have some pathogenic strains that are effectively resistant to all antibiotics"
The traditional test that hospital labs use relies on a substance called Mueller-Hinton broth, which enables many types of bacteria to grow. A sample from the infected patient is then pitted against different antibiotics so doctors can determine which works best—and how big a dose is needed. The process was standardized by the World Health Organization in 1961 so labs around the world could get comparable results. Back then, antibiotic treatments were relatively young, new classes of drugs were being discovered and superbugs weren’t widespread in hospitals and the world at large.
A big hurdle to developing new treatments is that there can be a big gap between a drug’s effectiveness in the laboratory, compared with the complexity of the human body. Several research labs have recently uncovered ways bacteria can elude the standard test, making invulnerable microbes appear susceptible to treatment. But in a new study in mice, scientists at the University of California-Santa Barbara may have hit on a way to make the test more accurate, and it involves using the most pedestrian of household items.
They took the standard antibiotic susceptibility test and added sodium bicarbonate, a chemical better known as baking soda. In addition to that small box containing it in the back of your refrigerator, it’s also found in human tissue—so researchers hypothesized that using it to test superbugs would better simulate how they behave in humans. “Now the bacteria is going, ‘I’m in the body, I need to fight,’" said Michael Mahan, a professor at UC Santa Barbara and senior author of the new study.
Mahan said more research is needed, but medicines did indeed behave differently when baking soda was added to the standard test. Other changes to the substance used to grow bacteria also affected how well the drugs seemed to work, showing some antibiotics to be more potent than previously known, and others less so.
But it’s still early days: The study doesn’t prove that the modified tests better predict what will work in the human body, independent experts said. Chambers, the infectious disease doctor, said the researchers are correct to point out the faults in the standard test and added, “there is room for improvement." But, he cautioned, “it is extremely difficult to make a link" between lab experiments and what works in patients.
For example, in the Santa Barbara study, mice were treated two hours after they were infected. Human patients would typically be infected for much longer before getting antibiotics, a factor that would affect how well the drugs work, Chambers said.
Despite the limitations, scientists need to continue exploring ways to make the tests more accurate as the threat of superbugs becomes more urgent, said Victor Nizet, a University of California-San Diego researcher who has worked on related research but was not involved in this paper. “We have some pathogenic strains that are effectively resistant to all antibiotics," Nizet said. Most of the time, the standard test works well, but when antibiotics don’t heal patients, “to say that, ‘We’re just going to stop at this one test and that’s it, that’s the only test we do for you,’ just seems ridiculous," Nizet said.
Melvin Weinstein, professor of pathology and laboratory medicine at Rutgers Robert Wood Johnson Medical School, said doctors have been debating the value of in vitro (outside the body) tests to predict an antibiotic’s potency for his entire 40-year career. “I think if we could have a better test method, we’d all say, ‘That’s great, let’s do it,’" he said.
The question for doctors and scientists is how to find a reliable method that’s simple enough to use in busy hospital labs. Baking soda may be a small step in that direction, but it will take a lot more research to prove that an alternative test works in the clinic. “This idea of trying to simulate the human milieu is a good one," said Weinstein. “I just think it’s going to be really hard to do." Bloomberg