P-737 Optimization of the multistep embryo transfer cycles in assisted reproductive technologies (ART) treatments

Abstract

Abstract Study question Decision of how many and which embryos to transfer in each cycle is determined by multiple factors, and the integration of patient-specific considerations is challenging. Summary answer We developed an embryo transfer strategy support tool (ET-SST) that finds the optimal way to conduct the embryo transfer process. What is known already In each transfer cycle, a decision of how many and which embryos to transfer is determined, based on embryo quality, risk of multiple pregnancies, maternal prognosis, cryopreservation factor, and the cost of each cycle. The parameters can be conflicting, high maternal receptivity might encourage single embryo transfer (SET) yet poor embryo quality might encourage multiple embryos transfer (MET) while co-considering cryopreservation, clinical risks, and future possible transfer. Current guidelines are not case-specific and only dictate a limit on the number of embryos to transfer based on maternal age solely, without regarding the heterogeneity of patients and several transfer cycles. Study design, size, duration We created a computational model integrating multiple parameters, assigning a specific embryo transfer option a value, based on the expected pregnancy rate. Assuming a given number of transferable embryos in a cycle, we calculate the expected value of a series of transfers (a strategy), using a recurrence relation. We applied dynamic programing that efficiently identifies the optimal strategy for the given cycle while considering maternal and other clinical parameters incorporated into the equation. Participants/materials, setting, methods The number of possible strategies is exponential with the number of embryos. There are 70 strategies with only four embryos viable to transfer, and the number increases to 4,550 with six embryos and over 100,000 with eight embryos. We created a recursive relation that describes the value of each strategy. Evaluating the different strategies is challenging even for modern computers. We apply dynamic programming to efficiently solve the recursion relation and identify the optimal strategy. Main results and the role of chance To assess the optimal transfer strategies, we developed the ET-SST and generated an online web application (https://etsst.researchsoftwarehosting.org/). For a given set of oocyte retrieval embryos, the ET-SST calculates all possible multistep embryo transfer strategies and, using the value function, finds the optimal strategy based on embryonic, maternal, and clinical user-input factors. (1) Embryo quality scores (using any method of choice). (2) Maternal prognosis score (endometrial receptivity, maternal age). (3) Multiple pregnancy score. (4) Cryopreservation factor (estimating the effects of embryo freeze-thaw procedure). (5) Transfer cycle cost. Each factor receives a value between 0 (non-favorable) and 1 (favorable). Based on these parameters, the ET-SST defines the optimal strategy that maximizes the value function. To avoid limitations on the assessment of the embryos' developmental potential and the patients' clinical prognosis, the ET-SST allows using all tools of choice for the user-input assignment. Limitations, reasons for caution This newly developed computerized model was not yet tested in a randomized trial. Wider implications of the findings Here we showed an algorithm that generates a case-specific strategy that can incorporate the clinical parameters that can be evaluated by any method that is preferred by the clinician. The modern evidence-based IVF clinic should include not only evidence-based clinical parameters but also algorithm-guided personalized decision making. Trial registration number exempted