An embryo cleavage pattern based on the relative blastomere size as a function of cell number for predicting implantation outcome.

Objective: To analyze whether the cleavage pattern redefined for all cleavage stages according to the relative blastomere size as a function of cell number has an additive value in predicting implantation potential of day-2 and day-3 embryos.
Design: Retrospective analysis of standard embryo morphologic parameters (cleavage rate and degree of fragmentation) supplemented by cleavage pattern findings of 347 implanted embryos compared with those of a matched control group of 307 embryos that failed to implant.
Setting: University-based tertiary medical center.
Patient(s): Two hundred and nine women with successful implantation and 181 controls matched for age and demographic parameters with failed implantation.
Intervention(s): In vitro fertilization (IVF), embryo assessment, and embryo transfer.
Main Outcome Measure(s): Determination of cleavage patterns in synchronized and unsynchronized cleaving embryos and correlations with implantation outcomes.
Result(s): Statistically significantly more embryos of the implanted group had good cleavage patterns compared with the failed implantation group (88% vs. 70%). A good cleavage pattern predicted implantation outcome even for nonsynchronized cleaving blastomeres at three, five, six, and seven cells (79% vs. 59%). Regression analysis demonstrated that adding cleavage pattern to the scoring system increased our ability to predict implantation in the training set; the area under the curve was the highest (0.707) as was the proportion of correct classification (>70%) when the cleavage pattern was assessed on both days 2 and 3.
Conclusion(s): When combined with measurements of the cleavage rate and degree of fragmentation, the cleavage pattern refines our ability to predict the likelihood of implantation, representing a definitive tool in the selection of top-quality embryos.
(Copyright © 2012 American Society for Reproductive Medicine. Published by Elsevier Inc. All rights reserved.)