A decision tree model for predicting live birth in FMR1 premutation carriers undergoing preimplantation genetic testing for monogenic/single gene defects.

ABSTRACT

RESEARCH QUESTION Can patient selection for successful preimplantation genetic testing for women who are fragile X (FMR1) premutation carriers be optimized using a decision tree analysis? This decision support tool enables a comprehensive study of a set of clinical parameters and the expected outcomes. DESIGN: A retrospective case-control study analysing the results of 264 fresh and 21 frozen preimplantation genetic testing for monogenic disorders/single gene defects (PGT-M) cycles in 64 FMR1 premutation carriers. Primary outcome was live birth per cycle start. Live birth rate was calculated for the start of the ovarian stimulation cycle. Fresh and frozen embryo transfers from the same cycle were included. RESULTS: The decision tree model showed that the number of cytosine guanine (CGG) repeats was only a moderate predictor for live birth, whereas an age younger than 36 years was the best predictor for live birth, followed by a collection of 14 or more oocytes. These findings were supported by the results of the logistic regression, which found that only age and oocyte number were significantly associated with live birth (P = 0.005 and 0.017, respectively). CONCLUSIONS: The number of CGG repeats is a relatively poor predictor for live birth in PGT-M cycles. FMR1 premutation carriers are no different from non-carriers. Age is the best identifier of live birth, followed by the number of retrieved oocytes.