New drugs for bad bugs: UF approach could bolster antibiotic arsenal

March 4, 2005

GAINESVILLE, Fla. — Call it a chemical crystal ball. A new approach to predict whether a drug in development is likely to work and which dose is best could get antibiotics to market faster and more cheaply, say University of Florida researchers.

In recent years, scientists worldwide have sounded the alarm: There simply aren’t enough drugs to combat bad bugs. Bacteria are increasingly adept at outwitting the traditional antibiotic arsenal.

Yet designing and testing new antibiotics can be a maddeningly slow and costly process — if pharmaceutical companies even bother, says Hartmut Derendorf, chairman of the department of pharmaceutics at the University of Florida College of Pharmacy. Many would rather invest in compounds aimed at patients with chronic conditions such as high cholesterol or diabetes, not in drugs designed to be used for a week or two and then stopped once an infection clears, he said.

Now UF researchers have devised a patent-pending method that combines testing of various drug concentrations at the site of infection with a series of laboratory analyses and mathematical models designed to streamline drug development. The method helps better determine which drugs are worth studying in people and at which dose, avoiding the typically lengthy and expensive trial-and-error approach that can take years.

“About one new antibiotic a year is approved,” said Derendorf, who will discuss the technique Saturday (March 5) at the annual meeting of the American Society for Clinical Pharmacology and Therapeutics in Orlando. “That’s certainly not enough. Even more worrisome, there are very few in the pipeline right now. Meanwhile, the requirements are getting longer and longer, and this is a huge dilemma with the recent discussion about Vioxx. That’s created some doubt in the approval procedure. I think we have to come to a reasonable expectation here in terms of the balance between benefits and potential harm. The worries I have right now are because of these unrealistic expectations, the requirements are going to be even higher, and it’s going to be harder and even more expensive to bring a new drug to market.”

About 70 percent of bacteria found in hospitals resist at least one of the drugs commonly used to treat the infections they cause, according to the Food and Drug Administration. The agency warns that unless problems are detected early and swift action taken to find substitute drugs, previously treatable diseases could again emerge in more virulent forms. Public health officials cite antibiotic resistance as a growing problem for a host of diseases, from childhood ear infections to malaria.

Last year, the FDA published a report calling attention to inefficiencies in the drug and medical product development process, urging changes to make the process “more predictable and less costly.” The latest estimates put the cost of bringing a new product to market at $1.6 billion to $1.8 billion.

UF researchers are working on an approach known as PK/PD, which combines principles of pharmacokinetics, or an analysis of drug concentrations in the body, and pharmacodynamics, their effect on bacteria or how a drug kills bacteria.

“In the past, blood samples were taken and the serum concentration of the drug was measured and that number was used to make dosing decisions,” said Derendorf, whose work is primarily funded by the pharmaceutical companies Pfizer and Sankyo. “That may not always be the right place to look. Most infections are not in the blood but in other sites of the body. Some of the recommendations we have may not be the optimal doses.”

UF researchers have developed a patent-pending technique called microdialysis that uses a small needle probe to measure how much of a drug actually ends up in the fluid surrounding the bacteria at sites of infection and are among the first in the country to test the method in people. These concentrations can differ widely from those found in the bloodstream, said Derendorf, who has published results from studies that evaluated the technique in people and animals with various infections. In the past, microbiologists would expose bacteria to certain concentrations of an antibiotic and then determine the minimum concentration that prevents bacterial growth. That number was taken and compared with concentrations of the drug in the blood, and from those two numbers a dosing decision is made.

“We feel that’s not the optimal way go,” he said. “It doesn’t give you the full story — it doesn’t tell you, for example, how quickly the bacteria are killed.”

So UF scientists developed a system of pumps they can use to expose bacteria to changing concentrations of an antibiotic, mimicking the concentration profile present in a patient at the actual site of an infection. They then measure how quickly the bacteria are killed or see if they regrow, and use mathematical modeling to estimate the optimal dose.

“Based on the results in the lab, then you can do a clinical study of what you think is going to work best,” Derendorf said, “and you’d find the best dose much faster than just by going by trial and error or by using some of the traditional ways.”

Consider one recent example: Derendorf led a series of laboratory experiments designed to evaluate an investigational, sustained-release form of a cephalosporin antibiotic. Ultimately the PK/PD approach showed that the difference in drug concentrations in the tissues arising from the standard form of the drug versus the sustained-release variety was so minimal, development of the new formulation was not warranted.

“The company that was interested in that decided not to continue that project any more and made that decision much sooner probably than they would have in earlier days,” said Derendorf, who published the findings in the Journal of Clinical Pharmacology. “Using the information early on to make a ‘no-go’ decision for a product so you don’t do a lot of other experiments to study a compound that later will be dropped — that alone saves a lot of money.”

UF researchers say they will continue to apply the screening approach to other drugs in various situations and also will seek to develop better ways of determining how frequently and at what dose a drug should be given to minimize the development of resistance. They recently collaborated, for example, with NASA to analyze blood and tissue concentrations of an antibiotic in people living for a few days in a simulated zero-gravity atmosphere.

“This approach is not just limited to anti-infectives,” Derendorf added. “We can expand it to other classes of drugs. It may be useful to answer many, many different questions.”

The PK/PD approach aims to ensure that doses tested in clinical trials are adequate and likely to yield good results, said Dr. William Craig, a professor of medicine at the University of Wisconsin, adding that until now, “many, many antibiotics that initially came out have either had their dosages changed because they were excessive or had them increased because they weren’t enough.”