Beyond Embryo Quality: Endometrial Factors in 66 Cryopreserved Embryo Transfers

Abstract

Background: Cryopreserved embryo transfer now accounts for the majority of transfers in many programs, and embryo selection using preimplantation genetic testing for aneuploidy (PGT-A) has reduced, but not eliminated, implantation failure. 1,4 When implantation fails despite the transfer of euploid embryos, attention turns to the endometrium – its receptivity timing, assessed by Endometrial Receptivity Analysis (ERA), 2,3,5 and its microbial composition, assessed by endometrial microbiome metagenomic analysis (EMMA).6,8 Both tests are offered routinely in this program, but their yield and their relationship to each other have not previously been audited.
Methods: Retrospective single-center cohort study of 66 cryo ET cycles performed in 35 patients between February 2024 and August 2026, using the clinic cycle register together with data from 40 ERA and 36 EMMA reports (Igenomix, processed at Avrupa Laboratories). Donor status was assigned per transfer by matching each transfer to its originating oocyte pick-up cycle by cryopreservation date rather than by patient-level flag. Pregnancy outcome was classified as clinical, biochemical, negative, or pending. ERA recommendations were recorded as hours of progesterone administration and compared with the biopsy protocol. EMMA reports were parsed for all panel organisms. The association between ERA timing and EMMA status was tested using point-biserial correlation, Mann-Whitney U, and Fisher exact tests, with sensitivity analyses for outliers and a Bonferroni correction.
Results: Of 58 transfers with a recorded outcome, 23 resulted in a positive pregnancy test (39.7%), comprising 14 clinical (24.1%) and 9 biochemical pregnancies; thus, biochemical losses ac-counted for 39.1% of all positive tests. Donor oocytes were used in 37 of 66 transfers (56.1%), and 53 transfers (80.3%) used PGT-A euploid embryos. Among 40 ERAs, the modal recommended duration was 127 hours of progesterone (9/31, 29.0%); however, in all 25 receptive reports that carried a numeric hour, the recommendation reproduced the biopsy protocol exactly, so the modal value reflects clinic practice rather than a biological optimum. Of the 35 women with an interpretable result, 8 (22.9%) had a displaced window of implantation. Five of 40 reports (12.5%) were technical failures. Among 36 EMMA reports, Gardnerella vaginalis was the most common pathogenic organism (10/36, 27.8%); Bifidobacterium spp was detect-ed more often (33.3%) but is not classed as pathogenic. In the 29 women with both tests, no association was found between ERA timing and EMMA findings (all p ≥ 0.38 after sensitivity analysis and correction for multiple testing).
Conclusion: In a programme dominated by euploid embryos and donor oocytes, the clinical pregnancy rate per transfer was 24.1%, and 39.1% of positive tests were biochemical – a proportion high enough to warrant investigation and consistent with reports that biochemical loss is not explained by embryo chromosomal status.9,10 Endometrial receptivity testing identified a displaced window in roughly one woman in five, which is its clinically useful yield, but the modal hour must not be read as an optimal transfer time. ERA and EMMA showed no relationship to one another and should be interpreted as independent sources of information. Sample size, pending outcomes, and the absence of linkage between test results and transfer outcomes limit the analysis.