Personalized Medicine Model Builder and Evaluation for Diabetes Treatments

Personalized Medicine Model Builder and Evaluation for Diabetes Treatments

None

None

In medical practice, when more than one treatment option is viable, there is little systematic use of individual patient characteristics to estimate which treatment option is most likely to result in a better outcome for the patient. For instance, in one study, some patients may have better outcomes on alogliptin with metformin than on glipizide with metformin (and vice-versa) and in another study, some patients may have better outcomes on HOE901-U300 than on Lantus (and vice-versa).
When making treatment allocation decisions such as deciding whether to give a patient glipizide or alogliptin (i.e. drugs in the studies we are applying for), practitioners often end up selecting treatments based on their own personal experiences or the experiences of their peers. If they are up-to-date with the literature or published RCTs, they may select the treatment which studies have found to be superior on average. Each of these approaches can sometimes lead to improved outcomes, but each also can be badly flawed. For example, in a variety of clinical settings, “craft lore” has been demonstrated to perform poorly, especially when compared to even very simple statistical models (Dawes,1979). It follows that each of these “business-as-usual" treatment allocation procedures can in principle be improved if patient characteristics, especially those related to how well an intervention performs, are taken into account.
Therefore, when deciding which of the two drugs to give a patient, It would be valuable to have a means of sorting patients using a statistical model that includes patient characteristics. It would also be valuable to estimate how clinically impactful the model will be when it is used to determine treatments for future patients. This type of system is not available presently and it would be of tremendous use to clinicians.
Generally our goal is to (1) create a sorting rule (e.g., a very simple one would be if BMI < 30 => Drug A otherwise if Age > 45 => Drug A otherwise Drug B) and then (2) report if this rule provides better clinical outcomes than “best” allocation (giving all future patients the treatment that performed best on average) and random sorting and (3) report the degree of statistical significance. Our goal can be attained by analyzing data from randomized clinical trials with (1) large sample sizes, (2) a lot of patient characteristic information is collected and (3) the treatments are drug A vs drug B (or placebo). Both studies we are applying for fit this description.
We are focusing our efforts now on Diabetes since there are a large number of such clinical trials investigating this disease. We will be broadening our scope to other diseases of worldwide importance in the future. Thus, currently there are 29.1 million people with Diabetes who would potentially benefit from our research.
We plan to publish results from our models in a top medical journal.

[{ "PostingID": 3512, "Title": "SANOFI-EFC11628", "Description": "6-Month, Multicenter, Randomized, Open-label, Parallel-group Study Comparing the Efficacy and Safety of a New Formulation of Insulin Glargine and Lantus® both plus Mealtime Insulin in Patients with Type 2 Diabetes Mellitus with a 6-Month Safety Extension Period

Medicine: Insulin Glargin 300 U/ml, Condition: Diabetes Mellitus Type 2, Phase: 3, Clinical Study ID: EFC11628, Sponsor: Sanofi" },{ "PostingID": 3929, "Title": "TAKEDA-SYR-322_305", "Description": "A Multicenter, Randomized, Double-Blind, Active-Controlled Study to Evaluate the Durability of the Efficacy and Safety of Alogliptin Compared to Glipizide When Used in Combination With Metformin in Subjects With Type 2 Diabetes

Medicine: alogliptin, Condition: Type 2 Diabetes Mellitus, Phase: 3, Clinical Study ID: SYR-322_305, Sponsor: Takeda" }]

The publication citation will be added after the research is published.