Mapping the way to successful euploid frozen embryo transfer: a prospective pilot study of uterine elastography

Does the objective and quantitative assessment of uterine tissue stiffness via ultrasound shear wave elastography (SWE) predict the outcome after single euploid frozen embryo transfer (FET)?

Uterine SWE data might be predictive of clinical pregnancy in good prognosis patients undergoing single euploid FET.

Few prior studies have evaluated the usefulness of strain uterine elastography in assessing the likelihood of conception in an infertile patient population following IUI or FET. These studies suggest that elastography parameters may be predictive of pregnancy following ART treatments. However, these studies are limited based on their use of strain uterine elastography, which provides a relative measurement between two areas of interest. No prior study has evaluated the more robust technology of SWE, which provides quantitative and objective measurements, on likelihood of conception following infertility treatment.

This is a prospective pilot study of 153 patients with no known uterine pathology undergoing single euploid FET at a university-affiliated center between June 2022 and December 2022.

Patients scheduled for upcoming single euploid blastocyst FET during the study period were evaluated for study participation. On the day prior to embryo transfer, SWE assessment of 15 regions of interest of the sagittal uterus measuring 6 mm in diameter was performed on included participants. Regions of interest included the cervix, anterior/posterior distal myometrium, anterior/posterior lower uterine segment, fundus, endometrium, and the anterior myometrial/endometrial interface. The elastography data points generated were collected in a standardized fashion by a single investigator blinded to patient outcome. The elasticity parameters analyzed included the mean, SD, minimum, and maximum values in kilopascals and meters per second. These data were assessed for relationship to clinical pregnancy, defined as discharge from fertility care at 9 weeks' gestation.

There were 22 uterine elastography parameters that showed a significant relationship to clinical pregnancy with P < 0.10 using univariant analyses. Regions of interest that were predictive of clinical pregnancy with P < 0.10 included the cervix, the anterior myometrium, the posterior myometrium, the fundus, and the anterior myometrial/endometrial interface. Increased mean (stiffer elastography metrics) in the posterior myometrium and anterior myometrial/endometrial interface was associated with increased likelihood of clinical pregnancy in comparison to FETs that did not result in clinical pregnancy. These values were then subjected to a multivariable logistic regression model to estimate areas under the curves and predictive values with and without clinical parameters. Using elastography data only, the model for predicting clinical pregnancy resulted in a significant AUC of 0.826 (95% CI: 0.748, 0.904). Next, an additional model was generated using both elastography data and patient demographic and cycle characteristics, which also had a significant AUC of 0.923 (95% CI: 0.874, 0.973). Receiver operating characteristic curves were compared to demonstrate that SWE and clinical parameters together provide a significantly better predictive value compared with just clinical parameters alone to forecast clinical pregnancy. We then ran a 5× cross-validation, confirming that the system was robust.

There are several limitations to this study. The model developed is predictive of clinical pregnancy in a good prognosis patient population with high-quality euploid embryos, thereby limiting the generalizability of the model in alternative patient populations. Further work is needed to validate our findings in larger patient populations. Additionally, SWE data may be limited by patient parameters beyond the sonographer's control.

Quantitative assessment of tissue elasticity using SWE of a morphologically normal uterus, assessed the day prior to single euploid FET, was found to result in a model predictive of clinical pregnancy in a good prognosis population.

Funding for this study was provided by IVIRMA Global. There are no conflicts of interest for any of the authors.

NCT05397912.