A new algorithm may take the guesswork out of medicating patients with cancer, bacterial infections, organ transplants, and other conditions that require very precise drug dosing.
Why it matters:
Individual differences can alter patient response to medications. Metabolism, body type, ethnicity, other illnesses, and genetics can play a role in how patients respond to drug treatment.
“When people talk about using genomics for precision, or personalized, medicine, that’s a very laudable goal,” said senior author Dean Ho, professor of oral biology and medicine at UCLA. “But I am more than just my genes. My body is going to be affected by things that have happened to me in the past.”
The nitty gritty:
The researchers call their method parabolic personalized dosing, or PPD. They gave patients medication and then observed the dosages which brought positive responses. Dosages were plotted along the X axis of a grid, with the patient’s best response, or output, placed on the Y axis. Where these two points met formed the beginning of a parabola. Then the dosage could be reduced or increased based on how much medication was in the patient’s blood, with the successful doses added to the parabola. The researchers called the parabola “a robust map that identifies drug doses (inputs) that ensure that a patient will stay in a target range.”
In a small pilot trial of eight patients who’d had liver transplants, PPD kept levels of an immunosuppressive drug more constant in patients’ blood than did physician-directed dosing. The findings appear in Science Translational Medicine.
But keep in mind:
Establishing a patient’s parabola still requires administering drugs and then observing the response, something doctors already do. “This allows us to make a better guess,” said first author Dr. Ali Zarrinpar, a transplant surgeon at UCLA. “When you give a drug treatment and you see a response, you know what amount of medications will provide that response.”
The bottom line:
In the age of big data, algorithms could play a helpful role in integrating a lot of patient information to make dosing decisions.
