13
August
2018
|
14:21
Europe/Amsterdam

AI for personalised medicine

Dr Chow (2nd from left) together with members of the team that developed QPOP (from left:) Prof Chng Wee Joo, Director of the National University Cancer Institute, Singapore, and Senior Principal Investigator at CSI; Dr Masturah Rashid, recent PhD graduate from NUS and first author of the study; and Prof Dean Ho, Director of SINAPSE 

NUS researchers have discovered a way to link artificial intelligence (AI) to personalisation of medicine. A multidisciplinary research team has developed an AI technology platform that can potentially change the way drug combinations are designed and allow doctors to more quickly determine the most effective drug combination for patients.

Existing techniques in the design of drug combinations usually involve testing arbitrary combinations of commonly used drugs or incorporating new targeted therapies into established drug combinations.

The research team, comprising clinicians, engineers and molecular biologists from NUS, applied their platform to a type of blood cancer — drug resistant multiple myeloma. Currently, Bortezomib-containing drug combinations are used as the first and second-line treatment for this cancer. Most patients inevitably become resistant to these drugs, however, and new combinations need to be quickly established. Although some newer combinations have been proven to be effective for some patients, rapid determination of an optimal personalised treatment for specific patients from the infinite range of possible drug combinations remains a challenge.

The newly-developed AI platform, called Quadratic Phenotypic Optimisation Platform (QPOP), aims to speed up drug combination design and to identify the most effective drug combinations for individual patients using small experimental data sets. Using a small amount of blood or bone marrow sample, the platform is able to map the drug response that a set of drug combinations will have on a specific patient’s cancer cells.

The efficiency of this platform in utilising small experimental data sets enables the identification of optimal drug combinations in a timely and cost-efficient manner, which marks a big leap forward in the field of personalised medicine.

Using QPOP, the researchers were able to identify a series of effective drug combinations from an initial pool of 114 FDA-approved drugs, including one completely novel and unexpected combination. The performance of this novel combination was validated against 13 patient samples and found to outperform the standard of care regimen for relapse myeloma. The team was also able to fine-tune the dosage ratios of the novel combination for optimal effectiveness using the platform.

When tested against two other clinically used drug combinations for four patients, the novel combination was found to be the most effective option for two myeloma patients. Although the novel combination was not the most effective for all four patients, the researchers also found that QPOP was able to match the best drug combination to each patient.

Dr Edward Kai-Hua Chow, Principal Investigator at the Cancer Science Institute of Singapore (CSI) at NUS, who led the study, said that QPOP revolutionises the way drug combinations are designed and represents a key area in healthcare that can be transformed with AI. “The efficiency of this platform in utilising small experimental data sets enables the identification of optimal drug combinations in a timely and cost-efficient manner, which marks a big leap forward in the field of personalised medicine,” he added.

The team is currently working towards translating this work. They plan to recruit patients for prospective clinical trials related to QPOP and other artificial intelligence platforms in 2019, as well as expand the application of the platform into other disease areas.

The Singapore Institute for Neurotechnology (SINAPSE) at NUS and the Biomedical Institute for Global Health Research and Technology at NUS were also involved in the development of QPOP.

The research was conducted in collaboration with the Agency for Science, Technology and Research, the National University Cancer Institute of Singapore and the University of California, Los Angeles and the findings were published in Science Translational Medicine in August.