Genetics, gender, diet, and other factors affect a drug’s effectiveness in a patient’s body, so UCSF pharmacists are building computer models to bring precision dosing into the mainstream.
Janel Boyle, PharmD, PhD (above), and Rada Savic, PhD, solve big problems for tiny patients, like the 6-month-old boy with leukemia who was about to undergo a bone marrow transplant. It was the UCSF pharmacists’ job to figure out the right dose of a drug to enhance his stem cell uptake. Too much and he might suffer irreversible brain toxicity; too little and the transplant might fail. The challenge in such cases? Standard dosing guidelines are based on trials in adults, not babies.
“The goal isn’t just that patients survive,” Boyle says, “but also making sure they ... don’t pay a lifelong consequence.” Based on computer models that Savic and Boyle developed, the baby received half the recommended dose to account for his age, weight, and kidney status. The result was a successful transplant with no toxic aftermath.
Adults face dosing problems too. It’s estimated that between 30 and 40 percent of the drugs people take don’t actually help. That’s because for every drug, explains B. Joseph Guglielmo, PharmD, resident alumnus and Dean of the UCSF School of Pharmacy, “there is a dose that is subtherapeutic, one that is therapeutic and not toxic, one that is therapeutic but toxic, and one that is flat-out toxic.”
Conventional dosing formulas are based mainly on age and weight. But many other factors – including genetics, gender, lifestyle, and diet – can affect how well a drug works in a given patient’s body. By contrast, more individualized dosing boosts drugs’ effectiveness while reducing or eliminating side effects.
To bring precision dosing into the mainstream, UCSF researchers like Savic and Boyle are building computer models based on real-world data. “It could be really personalized – how much you take, how frequently, and for how long,” says Savic, a specialist in pharmacometrics and precision dosing who has developed dosing models for drugs to treat diabetes, tuberculosis (TB), malaria, HIV, and leukemia in vulnerable populations.
It can take years to gather enough data to develop a model and then validate its complex calculations in a clinical trial. But some models – such as one Savic developed for a TB drug – have already been added to US Food and Drug Administration (FDA) advisories. And a spin-off from Savic’s lab, InsightRX, is now putting models into a user-friendly format that busy pharmacists and clinicians will find easy to use someday soon.