Prediction models and peptide profiling in the assessment of the implant ation potential of early embryos

Summary

In IVF/ICSI treatments, two or more embryos are often transferred to maximize the chance of success, which brings results in a high incidence of multiple pregnancy. Multiple pregnancies are associated with an increased incidence of mortality, a higher percentage of birth defects, increased incidence of low birth weight and premature deliveries. Performing single embryo transfers is the only way to avoid multiples pregnancies. This puts embryo selection to a challenge: choosing the embryo with the highest implantation potential so that the pregnancy rate remains high and the multiple pregnancy rate drops to a minimum.

Predicting outcome based on embryo selection based on assessment of development and morphology has limited value. This can be explained by a substantial intra- and inter-observer variability in morphological grading. Extra, novel non-invasive markers relevant to implantation potential are needed to improve the selection of the embryo with the highest implantation potential.

Therefore in the first section of this thesis, potential peptide biomarkers for oocyte/embryo viability were screened. The second section focused on the prediction model of embryo implantation potential using computer assisted scoring system. In the last section, the impact of clinical characteristics of patients on outcome was investigated in order to optimize patient counselling and to identify a reference population to allow benchmarking between clinics.

As a proof of principle, the peptide patterns were found to be promising in the discrimination of oocyte viability, regarding fertilization outcome, embryo development, and implantation success. Predictive models were constructed based on parameters measured with computer assisted scoring system, including the number of blastomeres, size difference between blastomeres, the percentage of fragmentation, and total cytoplasmic volume, showing a higher accuracy rate than models based on the standard morphologic assessments. The last section revealed that the exclusion of presence of tubal pathology in patients resulted in higher pregnancy rate and live birth rate, and a new reference population was suggested.

This project demonstrates the proof of principle of the use of follicular fluid biomarkers as an extra tool for embryo selection: prediction models for the outcome in a human IVF/ICSI program can be optimized using morphometric data and clinical parameters.