Deriving personalized and optimized estrogen dosage profiles for optimal endometrium thickness in in vitro fertilization patients

Objective In vitro fertilization (IVF) is an infertility treatment that inserts fertilized egg cells into the uterus of the patient. Before the transfer, estrogen dosing is used to thicken the endometrium so that the embryo can attach to the uterus lining. This study aims to develop a mathematical model that personalizes and optimizes individual patient estrogen dosage profiles to achieve optimal endometrial thickness (approximately 8 mm) during the endometrial preparation phase of IVF. Design Retrospective data on estrogen dosages vs. endometrial thickness was obtained. A physics-based model was developed for each patient by estimating patient-specific and independent model parameters. Nonlinear programming optimization techniques were used for parameter estimation as well as to derive an optimal dosage profile for each patient. Subjects 47 patients who received IVF treatment from Akanksha Hospital provided endometrium thickness data while receiving estrogen dosing. Exposure Patients’ endometrial growth data was used to estimate individual growth rates. Each patient was then exposed, in silico, to two estrogen dosing optimization models: an intermittent dosing regimen with a lower bound of 0 mg/day and a daily dosing regimen with a lower bound of 0.5 mg/day. Main Outcome Measures Coefficients for model fit, the total estrogen dose required to reach an endometrial thickness of 8 mm, and the percentage reduction in estrogen needed under each dosing optimization model were determined for each patient. Results 41 out of 47 patients could be modeled with a sufficient coefficient of determination. 39 out of the 41 patients were optimized to 8 mm while the remaining 2 out of 41 were optimized to their observed final thickness. All patients showed a reduction in estrogen under the optimization model. On average, patients had a larger percentage of estrogen reduction under the optimization with a lower bound of 0 mg/day than under the optimization with a lower bound of 0.5 mg/day. Conclusion Overall, the results show evidence that the model reduces the amount of estrogen needed to reach optimal thickness, with greater reductions under an intermittent dosing regimen. In the future, this model can be used for patients actively undergoing an IVF cycle by using the first two-day hormone and endometrial thickness data to personalize and optimize the estrogen dosing.