Is biochemical pregnancy loss (BPL) associated with embryo or endometrium? A multicentre retrospective study with+7000 cases

Abstract Study question Is biochemical pregnancy loss rate (BPLR) associated with alterations within the embryo or the endometrium? Summary answer The BPLR was similar in frozen embryo transfers among groups: euploid embryo (PGT-A group), adjusted by endometrial receptivity assay (ERA group), ERA+ PGT-A group, and controls. What is known already The etiology of BPL is not well known, but the most likely causes hypothesized seem to be either a low-quality embryo or impaired endometrial receptivity. Different studies showed that factors such as age, ovarian stimulation, the oocyte and embryo quality and endometrium receptivity may have something to do with the occurrence of biochemical pregnancy loss post-assisted reproduction treatments. However, other studies have shown that BPLR does not increase as a function of age or seems independent of embryo’s development stage or embryonic chromosomal constitution. No studies have evaluated nor controlled specifically the uterine factor with tools as ERA. Study design, size, duration Retrospective cohort study between January 2013 and January 2021, on patients undergoing 7491 frozen embryo transfers, comparing BPLR per pregnancy among groups of patients with optimized embryo and endometrial quality (ERA+ PGT-A, n = 169), optimized embryo quality (PGT-A, n = 2670), optimized endometrium quality (ERA=61), and controls (neither ERA nor PGT-A, n = 4591) from own oocytes, from two University-affiliated centers. We defined BPL as a pregnancy diagnosed only by serum beta-hCG>10UI/L without resulting in a clinical pregnancy. Participants/materials, setting, methods BPLR and other variables including age, body mass index, number of previous IVF, days of endometrial preparation, type of cycle (natural or with hormonal replacement), endometrial thickness, fresh sperm concentration and fresh sperm volume were evaluated among groups using chi-squared test, for univariate analysis. Main results and the role of chance Mean women’s age was 35.95 years 95%CI (35.86 ,36.04), BMI 22.52kg/m2 95%CI (22.45, 22.59), endometrial preparation length 17.96 days 95%CI (17.78, 18.14), 29.8% of them on natural cycle, and 70.25% with hormonal therapy replacement. Donor sperm was used on 7.8% of the treatments, and embryo quality was comparable between groups of study, with an overall A rating of the embryos 24.9%, B 50.95%, C 23.37% based on Spanish ASEBIR society criteria. Mean sperm characteristics were: volume 2.79 ml 95%CI (2.74-2.85), concentration 38.78 mill/ml 95%CI (33.77-39.39), on the raw ejaculate, and after preparation, concentration varied to 7.80mill/ml 95%CI(6.76-7.39) On the univariate analysis, overall BPLR was 8.06% 95%CI (7.45-8.70). Control group presented 7.65% 95%CI (6.89-8.46), with ERA group (optimized endometrial factor) showing 8.33% 95%CI (2.76-18.39), PGT-A group (optimized embryonic factor) 8.82% 95%CI (7.77-8.96), and PGT-A+ERA 7.14% 95%CI (3.75-12.14) without statistically significant differences. Limitations, reasons for caution As this was a retrospective cohort study, we are exposed to several biases that have been tried to control by means of inclusion/exclusion criteria and proper statistical adjustment. Subsequently, results should be interpreted with caution, and a clear cause-effect relationship cannot be assumed. Wider implications of the findings Our results indicate that with PGT-A, ERA test or both it is not possible to modify the BPLR. A different mechanism no associated with chromosomal structure or transcriptomic expression of the endometrium could be studied as cause of BPL. More studies are needed for explore this reproductive issue. Trial registration number not applicable