Combination therapy is a major strategy to circumvent the onset of treatment resistance in cancer patients. However, it can be very challenging. For M drugs with N doses each, there will be N to the Mth power of combinations. How to home in the optimal drug-dose combination from such a large number of choices for a specific subject is a non-trivial task. Professor Chih-Ming Ho and colleagues developed an AI based phenotypic response surface platform. With a few drug-dose combinatorial treatments and the measured efficacy/toxicity responses from the subject, we can come up an parabolic shaped map, which guides us to reach the optimized drug-dose combination on the response surface.

In the paper entitled “Optimizing drug combinations against multiple myeloma using a quadratic phenotypic optimization platform (QPOP)” published in Science Translational Medicine, we used a small amount of blood or bone marrow sample to map the drug response for a specific patient’s multiple myeloma cancer cells over a xenograft mouse model. We can optimized the dosages of each drug within a given combination while minimizing overall toxicity.

Please visit Prof. Ho’s new website: Ho Lab AI-Personalized Medicine.

  •  
  •  
  •  

 UCLA Samueli Mechanical and Aerospace Engineering

×