Home / Resources / Articles / Precision Medicine Models To Optimize Diabetes Treatment

Precision Medicine Models To Optimize Diabetes Treatment

Dec 1, 2020
Editor: Steve Freed, R.PH., CDE

Author: Shana Indawala, PharmD Candidate, University of South Florida Health, Taneja College of Pharmacy

A perspective on the selection of the optimal type 2 diabetes treatment based on precision medicine-based strategies.  

Although glucagon-like peptide 1 receptor agonist (GLP-1RA) and sodium-glucose cotransporter-2 inhibitors (SGLT2I) have established benefits in cardiovascular disease patients, heart failure, and chronic kidney disease, further data is limited on the benefit of each antidiabetic class in the remaining population. Models for using precision medicine in T2DM exist for monogenic diabetes but require expensive genetic testing to determine the specific etiology to aid in treatment selection. Due to this gap, approaches are sought to characterize patients beyond the T2DM phenotype and use heterogeneity to optimize the selection of diabetic treatment. Due to the different mechanisms of actions available, and each class of antidiabetic having varying glucose-lowering response, the use of precision medicine for T2DM treatment becomes ideal.  

This study aims to determine the selection of optimal T2DM treatment based on differences in drug effects. The heterogeneity of treatment effect (HTE) is utilized to choose the treatment selection since it is plausible that patients with different underlying pathophysiology will have varying responses to the other drug classes. Studies of HTE in T2DM, along with existing clinical and trial data, are utilized to develop and test precision medicine-based strategies with the intent to optimize treatment. This study does not aim towards identifying patients who will or will not respond to classes of medication, but instead seeks to identify patient populations that are more likely to have a more significant relative benefit with the use of a particular class. The U.K.’s Clinical Practice Research Datalink was utilized as an extensive, anonymized clinical health record database to provide patient demographics, clinical features, and laboratory tests. The total data points from all the drug classes were used for a discovery analysis in which drug-by-marker interactions were identified. To reduce bias, confounders and statistical adjustments within previous studies were identified. Studies were confirmed using a second step of external validation.  

U.K’s primary care observational data set was used to determine the difference in the benefit of sulfonylurea and thiazolidinedione in patients. Males without obesity, classified as having a BMI < 30, were more likely to have a more significant glucose reduction response with sulfonylurea than thiazolidinedione treatment. On the other hand, females with obesity, classified as having a BMI ≥ 30, were more likely to receive a more significant glucose reduction response with thiazolidinedione use than sulfonylureas. Dipeptidyl peptidase-4 inhibitors (DPP-4I) and GLP-1RA were compared next. The Predicting Response to Incretin Based Agents study showcased that higher insulin resistance was associated with a decreased response to glucose reduction in non-insulin-treated patients. Within this subsect of patients, patients with obesity and high triglycerides had less than half the response of DPP-4I than patients that without obesity and with low triglycerides. There was no evidence of an association between insulin resistance and response to glucose reduction in non-insulin-treated patients for the use of GLP-1RA.  

The use of sodium-glucose cotransport2 inhibitors (SGLT2I) and DPP-4I in the patient were also assessed. Patients with a higher baseline HbA1c may benefit from SGLT2I compared to DPP-4I or sulfonylurea treatment. The renal function of patients on SGLT2I should be considered. Patients with eGFR > 90 mL/min/1.73m2 may achieve a greater response to glucose reduction compared to patients with an eGFR of 60-90 mL/min/1.73m2. Patients with an eGFR < 60 mL/min/1.73m2 may experience reduced efficacy with SGLT2I, but patients on DPP-4I with reduced eGFRs are more likely to maintain efficacy.  

The provided evidence through studies and the U.K. database allows precision medicine implementation through two approaches: a subtype approach and an individualized prediction approach. The subtype approach will enable patients with T2DM to be classified based on their pathophysiology of clinical, genetic, phenotypic, or biomarker traits. The individualized prediction approach focuses on the markers in the underlying pathophysiology to predict a patient’s treatment response. Due to the association between a low BMI and lower insulin resistance, the use of DPP-4I may be beneficial in patients of Asian ethnicities. Additionally, patients with lower eGFR may benefit from DPP-4I compared to SGLT2I. Further studies on patients with microvascular and macrovascular complications should be conducted to determine nonglycemic endpoints.  

Practice Pearls: 

  • DPP-4I may be more beneficial in patients of Asian descent with a low BMI and lower insulin resistance.  
  • Patients with a reduced eGFR may achieve more benefit in diabetes management with DPP-4I.  
  • Patients with a reduced eGFR of < 60 mL/min/1.73m2 on an SGLT2I may experience a reduced response to glucose reduction.  


Dennis, John M. “Precision Medicine in Type 2 Diabetes: Using Individualized Prediction Models to Optimize Selection of Treatment.” Diabetes, vol. 69, no. 10, 2020, pp., 2075–2085., doi:10.2337/dbi20-0002. 

Kent, David M., et al. “The Predictive Approaches to Treatment Effect Heterogeneity (PATH) Statement.” Annals of Internal Medicine, vol. 172, no. 1, 2019, p. 35., doi:10.7326/m18-3667. 


Shana Indawala, PharmD Candidate, University of South Florida Health, Taneja College of Pharmacy