ROLE OF BLASTOCYST MORPHOLOGY IN PREDICTING CLINICAL OUTCOMES IN SINGLE FROZEN BLASTOCYST TRANSFERS

Abstract

Listen

Translate article

To assess the value of individual blastocyst morphological parameters in predicting clinical outcomes in single frozen blastocyst transfers. A retrospective cohort study in which, 1046 frozen embryo transfers (FETs), performed between June 2015 and June 2019, were analyzed. Only autologous, single, frozen blastocyst transfers with female age ≤ 38 years were included in the analysis. Those FET cycles were excluded in which, either, partially damaged blastocysts or blastocysts that failed to show any re-expansion post warming, were transferred. All frozen thawed blastocysts included in the analysis were originally graded prior to vitrification, on day 5 or 6 of culture, on the basis of morphology and individual scores were assigned to each blastocyst in terms of degree of expansion, inner cell mass (ICM) and trophectoderm (TE) cells using the Gardner and Schoolcraft scoring system. Multi-collinearity test and logistic regression analysis was used to study the relationship between blastocyst morphology and FET cycle outcomes. Univariate analysis showed that there is very little correlation between blastocyst morphology and the likelihood of implantation or that of a pregnancy loss, subsequently. Multivariate analysis showed that implantation rates were highest when a blastocyst with the following combination, having degree of expansion = 5, ICM = A and TE = A, was transferred. For all other combinations, little or no association was observed with both, implantation rates and pregnancy loss rates. Outcome prediction using logistic regression showed that grade 5 hatching blastocyst was the only statistically significant variable in predicting implantation. Prediction accuracy of 55.1% and 33.8% was observed for implantation and pregnancy loss, respectively. Our data shows that none of the parameters used to assess morphology were predictive of implantation for frozen blastocysts. The pregnancy loss rates were also found to be same irrespective of morphology or degree of expansion. Prediction accuracy of 33% for pregnancy loss showed that blastocyst morphology in no way governed the clinical likelihood of a pregnancy loss.