Late luteal progesterone as a marker of implantation success rather than its determinant in modified natural cycle single euploid frozen embryo transfer: a prospective multicenter study

In modified natural cycle frozen embryo transfers, β-hCG day progesterone is a strong predictor of ongoing pregnancy, while transfer-day progesterone does not provide useful prognostic information.

Summary and findings

This study evaluated the prognostic value of serum progesterone on the day of embryo transfer and the day of the β-hCG pregnancy test in patients undergoing single euploid frozen blastocyst transfer. A total of 207 patients were included, with ovulation triggered by recombinant hCG 6500 IU and vaginal micronized progesterone initiated two days later. The study found that β-hCG day progesterone was significantly higher in patients with ongoing pregnancy compared to those without.

Abstract

<title>Abstract</title> <p> <bold>Purpose</bold> To evaluate the prognostic value of serum progesterone measured on the day of embryo transfer and on the day of the β-hCG pregnancy test for reproductive outcomes in patients undergoing single euploid frozen blastocyst transfer in a modified natural cycle (mNC-FET). <bold>Methods</bold> Prospective nested cohort study conducted across six centres of the Instituto Bernabeu network (Alicante, Madrid, Albacete, Cartagena, Elche, and Mallorca) between 2023 and 2025. A total of 207 patients undergoing mNC-FET of a single PGT-A-screened euploid blastocyst were included. Ovulation was triggered with recombinant hCG 6500 IU and vaginal micronized progesterone (400 mg/night) was initiated two days later. Serum progesterone was measured by electrochemiluminescence immunoassay (Roche Cobas e411) on the day of embryo transfer (hCG + 7) and on the day of the β-hCG test (hCG + 16). The primary outcome was ongoing pregnancy, defined as a viable intrauterine pregnancy with cardiac activity at 10 weeks of gestation. ROC curve analysis and logistic regression were used to assess the discriminatory capacity of each measurement. <bold>Results</bold> The overall ongoing pregnancy rate was 109/207 (52.7%). Transfer-day progesterone did not differ significantly between patients with and without ongoing pregnancy [26.1 (IQR 20.6–32.0) vs 26.4 (IQR 22.0–31.1) ng/ml; p = 0.903] and showed no discriminatory capacity for any reproductive outcome (AUC 0.505, 95% CI 0.424–0.587). In contrast, β-hCG day progesterone was significantly higher in patients with ongoing pregnancy than in those without [34.4 (IQR 26.4–41.9) vs 12.8 (IQR 9.3–19.1) ng/ml; p < 0.001], with an AUC of 0.877 (95% CI 0.818–0.936) and an optimal threshold of 20.9 ng/ml (sensitivity 88.6%, specificity 80.3%). This association remained significant after multivariable adjustment (adjusted OR 1.15, 95% CI 1.10–1.20; p < 0.001). The luteal progesterone ratio (β-hCG day / transfer day) showed comparable discriminatory performance (AUC 0.871, 95% CI 0.805–0.937; sensitivity 96.1%, specificity 78.6%) without adding predictive value beyond the absolute β-hCG day measurement. <bold>Conclusions</bold> In mNC-FET with single euploid blastocyst transfer, progesterone on the day of embryo transfer is not predictive of reproductive outcomes, while β-hCG day progesterone is a strong and independent predictor of ongoing pregnancy. The comparable performance of the absolute β-hCG day value and the luteal ratio, combined with uniformly adequate transfer-day progesterone across all outcome groups, suggests that low late-luteal progesterone in this setting reflects a consequence of implantation failure - mediated by absent trophoblastic luteotrophic signalling - rather than its cause. These findings challenge the rationale for progesterone rescue supplementation in mNC cycles and provide the biological framework for future interventional studies in this population. </p>

Limitations

Not reported in abstract.

Elsewhere in the HCG (Human Chorionic Gonadotropin) corpus