Serum progesterone concentrations at frozen embryo transfer in women with adenomyosis and endometriosis: a retrospective cohort study

A total of 1,927 patients were included in the study, including 327 women with endometriosis (237 HRT, 90 mNC), 359 with adenomyosis (267 HRT, 92 mNC) and 1,344 serving as controls (Fig.  1 ). Within the endometriosis group, 300 women (91.7%) had endometriomas, 24 (7.3%) had DIE and 3 (1%) presented with only SUP. In the adenomyosis group, 351 women (97.8%) had diffuse adenomyosis, while 8 (2.2%) had adenomyomas. Baseline characteristics are shown in Table  1 , with protocol-stratified analyses provided in Supplementary Table 1. Women with endometriosis and adenomyosis had significantly lower AMH levels compared with controls ( p -value = 0.001 and p -value = 0.002, respectively), whereas BMI and smoking status were comparable across groups. Fig. 1 Study flowchart of patient selection and inclusion. Flowchart illustrating patient selection and inclusion in the study. Reasons for exclusion may overlap Study flowchart of patient selection and inclusion. Flowchart illustrating patient selection and inclusion in the study. Reasons for exclusion may overlap Table 1 Demographics and clinical characteristics Characteristics Controls ( n = 1344)        Endometriosis   ( n = 327) p -value Adenomyosis ( n = 359) p -value Age (years) 37.49±4.28 37.86±4.24 0.106 38.23±4.31 0.004 BMI (kg/m 2 ) 22.15±3.30 22.00±3.11 0.456 22.38±3.49 0.247 Smoking 154 (11.5%) 35 (10.7%) 0.699 49 (13.6%) 0.255 Main cause of infertility  Male infertility 345 (25.7%) 56 (17.1%) 0.001 75 (20.9%) 0.062  Tubal infertility 95 (7.1%) 25 (7.6%) 0.717 33 (9.2%) 0.175  Endometriosis 0 (0%) 327 (100%) - 103 (28.7%) -  Diminished ovarian reserve 109 (8.1%) 29 (8.7%) 0.655 56 (15.6%) <0.0001  Ovulatory 130 (9.7%) 7 (2.1%) <0.0001 29 (8.1%)  0.356  Recurrent pregnancy loss 92 (7.4%) 16 (4.9%) 0.198 24 (6.7%) 0.915  Multiple female factors 115 (8.6%) 74 (22.6%) <0.0001 65 (18.1%) <0.0001  Mixed 113 (8.4%) 56 (17.1%) <0.0001   41 (11.4%) 0.077   Unexplained infertility 591 (44.0%) 0 (0%)     - 112 (31.2%) <0.0001 Presence of adenomyosis 0 (0%) 103 (31.5%) - 359 (100%) -  Diffuse adenomyosis 0 (0%) 101 (30.9%) - 351 (97.8%) -  Adenomyoma 0 (0%) 2 (0.6%) - 8 (2.2%) - Presence of endometriosis 0 (0%) 327 (100%) - 103 (28.7%) -  Endometrioma 0 (0%) 300 (91.7%) - 89 (24.8%) -  DIE 0 (0%) 24 (7.3%) - 14 (3.9%) -  SUP 0 (0%) 3 (1.0%) - 0 (0%) - AMH (ng/mL) 2.97±2.67 2,43±2.32 0.001 2.48±2.28 0.002 Type of E2 treatment None 1 391 (29.1%) 90 (27.5%) 0.574 92 (25.6%) 0.196  Transdermal 50 (3.7%) 8 (2.4%) 0.259 0 (0%) <0.0001  Oral 772 (57.4%) 208 (60.9%) 0.042 267 (74.4%) <0.0001  Both 131 (9.7%) 21 (6.4%) 0.061 0 (0%) <0.0001 Endometrial thickness 8.41±1.40 8.51±1.41 0.248 8.42±1.40 0.904 Descriptive statistics were used to summarize patient characteristics, with continuous variables presented as mean ± standard deviation (SD) and categorical variables as counts and percentages. Comparisons between groups (endometriosis vs. controls and adenomyosis vs. controls) were performed using univariate regression models, specifically linear regression for continuous outcomes and logistic regression for categorical outcomes  DIE Deep infiltrating endometriosis, SUP Superficial peritoneal endometriosis P -values < 0.05 were considered statistically significant and are shown in bold 1 Only spontaneous modified cycles Demographics and clinical characteristics Descriptive statistics were used to summarize patient characteristics, with continuous variables presented as mean ± standard deviation (SD) and categorical variables as counts and percentages. Comparisons between groups (endometriosis vs. controls and adenomyosis vs. controls) were performed using univariate regression models, specifically linear regression for continuous outcomes and logistic regression for categorical outcomes DIE Deep infiltrating endometriosis, SUP Superficial peritoneal endometriosis P -values < 0.05 were considered statistically significant and are shown in bold 1 Only spontaneous modified cycles In HRT-FET cycles, median serum progesterone levels on the day of embryo transfer were comparable between women with endometriosis (13.9 ng/mL, interquartile range, IQR 10.9–17.7), adenomyosis (13.6 ng/mL, IQR 10.4–18.4), and unaffected controls (13.8 ng/mL, IQR 10.7–18.2). Linear regression analysis confirmed the absence of significant differences compared with controls (endometriosis: β = 0.02, 95% CI − 0.82 to 0.85, p -value = 0.968; adenomyosis: β = 0.14, 95% CI − 0.80 to 1.07, p -value = 0.773). Similarly, in mNC-FET cycles, progesterone concentrations were comparable across diagnostic groups, with median levels of 24.3 ng/mL (IQR 20.5–31.5) in women with endometriosis, 25.5 ng/mL (IQR 21.8–30.1) in adenomyosis, and 25.3 ng/mL (IQR 21.3–31.1) in controls. Adjusted linear models showed no significant association between disease status and progesterone levels (endometriosis: β = −0.66, 95% CI − 2.67 to 1.34, p  = 0.517; adenomyosis: β = 0.87, 95% CI − 1.29 to 3.04, p -value = 0.428). A multivariate linear regression analysis was conducted to examine the factors influencing progesterone levels on the day of FET in the overall study population. Given the significant differences in progesterone levels between HRT and mNC cycles, analyses were performed separately for the two protocols (Table  2 ), incorporating all the previously described variables. In both analyses, BMI emerged as the primary statistically significant variable influencing progesterone levels. Additionally, in the mNC analysis, estradiol levels on the day of FET also yielded a statistically significant result. Table 2 Multivariable linear regression analyzing the relationship between patient characteristics and the progesterone level on the day of FET in HRT cycles and in mNC cycles HRT-FET mNC-FET β 95% CI p -value β 95% CI p -value Age -0.028 -0.114/0.057 0.514 -0.048 -0.250/0.154 0.639 BMI -0.190 -0.283/-0.097 < 0.0001 -0.379 -0.623/-0.134 0.002 Estradiol levels -0.001 -0.004/0.002 0.396 0.007 0.001/0.014 0.023 AMH -0.033 -0.163/0.097 0.621 -0.137 -0.532/0.258 0.495 Endometrial thickness -0.150 -0.378/0.078 0.197 0.049 -0.469/0.567 0.852 Smoking 0.156 -0.782/1.095 0.744 0.838 -1.636/3.311 0.506 Male infertility 0.124 -0.628/0.877 0.746 -0.890 -2.555/0.774 0.294 Ovulatory infertility -0.499 -1.622/0.623 0.383 -1.401 -5.234/2.433 0.473 DOR 0.757 -0.375/1.889 0.190 -0.755 -3.276/1.767 0.557 Recurrent pregnancy loss -0.604 -1.912/0.703 0.364 -1.102 -3.921/1.716 0.443 Tubal infertility 0.792 -0.382/1.967 0.186 -0.589 -3.432/2.255 0.684 Endometriosis -0.100 -0.949/0.749 0.817 -0.801 -2.779/1.176 0.426 Adenomyosis 0.026 -0.812/0.864 0.952 1.453 -0.540/3.446 0.153 Transdermal vs. oral estradiol 1 0.655 -0.269/1.579 0.165 NA NA NA Live birth 0.044 -0.635/0.723 0.899 0.734 -0.886/2.354 0.374 Progesterone levels were treated as the dependent variable, while all the variables listed in the table were considered independent variables. The results for each variable are presented as the estimate, 95% confidence interval (CI), and p -value. P -values < 0.05 were considered statistically significant and are shown in bold DOR   Diminished ovarian reserve,  FET Frozen-thawed embryo transfer, HRT   Hormone-replacement therapy, mNC   Modified natural cycle 1 Women who did both oral and transdermal estradiol were included in the transdermal group Multivariable linear regression analyzing the relationship between patient characteristics and the progesterone level on the day of FET in HRT cycles and in mNC cycles Progesterone levels were treated as the dependent variable, while all the variables listed in the table were considered independent variables. The results for each variable are presented as the estimate, 95% confidence interval (CI), and p -value. P -values < 0.05 were considered statistically significant and are shown in bold DOR   Diminished ovarian reserve,  FET Frozen-thawed embryo transfer, HRT   Hormone-replacement therapy, mNC   Modified natural cycle 1 Women who did both oral and transdermal estradiol were included in the transdermal group Further regression analyses evaluated differences in progesterone levels between affected patients and healthy controls according to reproductive outcome. In HRT-FET cycles, among women achieving live birth, progesterone levels were similar between endometriosis and controls (β = −0.06, 95% CI − 1.44 to 1.32, p -value = 0.935), while a non-significant increase was observed in adenomyosis compared with controls (β = 1.22, 95% CI − 0.10 to 2.54, p -value = 0.069). Among women with no live birth, progesterone concentrations remained comparable between endometriosis and controls (β = 0.06, 95% CI − 0.99 to 1.10, p -value = 0.914) and between adenomyosis and controls (β = −0.70, 95% CI − 1.69 to 0.29, p -value = 0.166) (Table  3 ). Table 3 Progesterone levels on the day of FET in women undergoing HRT, categorized by outcome (live birth vs. no live birth) Mean ± SD Median (IQR) β 95% CI p -value Live birth = 450  All women = 450 14.79  ±  5.68 13.90 (10.80–18.20) - - -  Controls = 310 14.52  ±  5.67 13.75 (10.40–18.10) Ref Ref Ref  Endometriosis = 80 14.47  ±  5.28 13.90 (11.15–17.60) -0.06 −1.44–1.32 0.935  Adenomyosis = 94 15.74  ±  5.81 14.15 (11.80-19.47) 1.22 −0.10–2.54 0.069 No live birth = 921  All women = 921 14.59  ±  5.98 13.70 (10.60–18.00) - - -  Controls = 643 14.69  ±  5.93 13.80 (10.80-18.25) Ref Ref Ref  Endometriosis = 157 14.75  ±  6.22 13.70 (10.70–18.00) 0.06 −0.99–1.10 0.914  Adenomyosis = 173 13.99  ±  5.77 13.30 (9.81–17.30) −0.70 −1.69 − 0.29 0.166 Progesterone levels (ng/mL) are presented for each group as mean ± standard deviation (SD) and median (interquartile range, IQR). Linear regression analysis was used to compare progesterone levels between groups, with controls as the reference category. Results are presented as regression coefficients (β) with 95% confidence intervals (CI) and corresponding p -values FET   Frozen-thawed embryo transfer, HRT   Hormone-replacement therapy Progesterone levels on the day of FET in women undergoing HRT, categorized by outcome (live birth vs. no live birth) Progesterone levels (ng/mL) are presented for each group as mean ± standard deviation (SD) and median (interquartile range, IQR). Linear regression analysis was used to compare progesterone levels between groups, with controls as the reference category. Results are presented as regression coefficients (β) with 95% confidence intervals (CI) and corresponding p -values FET   Frozen-thawed embryo transfer, HRT   Hormone-replacement therapy In mNC-FET cycles, progesterone levels were also comparable between endometriosis and controls both among women achieving live birth (β = −0.85, 95% CI − 4.47 to 2.77, p -value = 0.644) and among patients with no live birth (β = −0.57, 95% CI − 3.00 to 1.85, p -value = 0.642). Likewise, no significant differences were detected between adenomyosis and controls among live birth cycles (β = 2.22, 95% CI − 1.66 to 6.09, p -value = 0.260) or no live birth cycles (β = 0.47, 95% CI − 1.80 to 2.73, p -value = 0.684) (Table  4 ). Table 4 Progesterone levels on the day of FET in women undergoing mNC, categorized by outcome (live birth vs. no live birth) Mean ± SD Median (IQR) β 95% CI p -value Live birth = 160  All women = 160 27.05  ±  8.56 26.60 (22.50-31.75) - - -  Controls = 117 26.90  ±  8.53 26.00 (22.40–32.10) Ref Ref Ref  Endometriosis = 25 26.05  ±  7.20 26.80 (21.90–30.60) −0.85 −4.47–2.77 0.644  Adenomyosis = 22 29.11  ±  7.86 27.65 (26.35–29.88) 2.22 −1.66–6.09 0.260 No live birth = 396  All women = 396 26.24  ±  8.88 24.85 (20.70–30.80) - - -  Controls = 274 26.28  ±  8.70 25.20 (20.75–30.58) Ref Ref Ref  Endometriosis = 65 25.70  ±  9.89 23.60 (20.20–32.00) −0.57 −3.00-1.85 0.642  Adenomyosis = 70 26.74  ±  8.19 24.65 (21.43–30.72) 0.47 −1.80–2.73 0.684 Progesterone levels (ng/mL) are presented for each group as mean ± standard deviation (SD) and median (interquartile range, IQR). Linear regression analysis was used to compare progesterone levels between groups, with controls as the reference category. Results are presented as regression coefficients (β) with 95% confidence intervals (CI) and corresponding p -values FET   Frozen-thawed embryo transfer, mNC   Modified natural cycle Progesterone levels on the day of FET in women undergoing mNC, categorized by outcome (live birth vs. no live birth) Progesterone levels (ng/mL) are presented for each group as mean ± standard deviation (SD) and median (interquartile range, IQR). Linear regression analysis was used to compare progesterone levels between groups, with controls as the reference category. Results are presented as regression coefficients (β) with 95% confidence intervals (CI) and corresponding p -values FET   Frozen-thawed embryo transfer, mNC   Modified natural cycle Additional regression analyses examined the association between reproductive outcome and progesterone levels within each diagnostic group. In HRT-FET cycles, progesterone concentrations did not differ between women achieving live birth and women with no live birth with endometriosis (β = −0.28, 95% CI − 1.88 to 1.33, p -value = 0.735) or among unaffected controls (β = −0.16, 95% CI − 0.95 to 0.63, p -value = 0.691). In contrast, women with adenomyosis who achieved live birth showed significantly higher progesterone levels compared with patients with no live birth (β = 1.76, 95% CI 0.30 to 3.22, p -value = 0.018) (Table  3 ; Fig.  2 ). Fig. 2 Violin plot of progesterone levels on the day of FET in women undergoing HRT, categorized by outcome (live birth vs. no live birth). Violin plots showing the distribution of serum progesterone levels on the day of FET across study groups. The central line represents the median, and the shaded area reflects data density. Data are presented as median and interquartile range (IQR). P -values (p) indicate the comparison between live birth and no live birth within each diagnostic group (adenomyosis, endometriosis, and controls). FET = frozen-thawed embryo transfer, HRT = hormone-replacement therapy Violin plot of progesterone levels on the day of FET in women undergoing HRT, categorized by outcome (live birth vs. no live birth). Violin plots showing the distribution of serum progesterone levels on the day of FET across study groups. The central line represents the median, and the shaded area reflects data density. Data are presented as median and interquartile range (IQR). P -values (p) indicate the comparison between live birth and no live birth within each diagnostic group (adenomyosis, endometriosis, and controls). FET = frozen-thawed embryo transfer, HRT = hormone-replacement therapy In mNC-FET cycles, progesterone levels were comparable between live birth and women with no live birth with endometriosis (β = 0.35, 95% CI − 3.97 to 4.66, p -value = 0.874) and among unaffected controls (β = 0.62, 95% CI − 1.26 to 2.50, p -value = 0.515), whereas no significant association was observed in adenomyosis (β = 2.37, 95% CI − 1.57 to 6.31, p -value = 0.235) (Table  4 ; Fig.  3 ). Fig. 3 Violin plot of progesterone levels on the day of FET in women undergoing mNC, categorized by outcome (live birth vs. no live birth). Violin plots showing the distribution of serum progesterone levels on the day of FET across study groups. The central line represents the median, and the shaded area reflects data density. Data are presented as median and interquartile range (IQR). P -values (p) indicate the comparison between live birth and no live birth within each diagnostic group (adenomyosis, endometriosis, and controls). FET = frozen-thawed embryo transfer, mNC = modified natural cycle Violin plot of progesterone levels on the day of FET in women undergoing mNC, categorized by outcome (live birth vs. no live birth). Violin plots showing the distribution of serum progesterone levels on the day of FET across study groups. The central line represents the median, and the shaded area reflects data density. Data are presented as median and interquartile range (IQR). P -values (p) indicate the comparison between live birth and no live birth within each diagnostic group (adenomyosis, endometriosis, and controls). FET = frozen-thawed embryo transfer, mNC = modified natural cycle Sensitivity analyses using log-transformed progesterone values confirmed the robustness of the main findings. No significant changes were observed in the direction, magnitude, or statistical significance of the associations. In HRT-FET cycles, no significant differences were found between diagnostic groups and controls, and within-group analyses confirmed the absence of associations between progesterone levels and reproductive outcomes in women with endometriosis and in controls. The previously observed association between progesterone levels and live birth in women with adenomyosis undergoing HRT-FET remained statistically significant after log transformation. Similar results were observed in mNC-FET cycles, with no significant associations across diagnostic groups or outcomes. To further evaluate the robustness of the findings, a sensitivity analysis using mutually exclusive diagnostic groups was conducted (Supplementary Table 2). In this analysis, progesterone levels were largely comparable across diagnostic categories. The previously observed significant difference within the adenomyosis group, namely higher progesterone levels in women achieving live birth compared with those not achieving live birth in HRT-FET cycles, was confirmed. In addition, women achieving live birth with adenomyosis showed higher progesterone levels compared with controls. A subgroup analysis was performed restricting the population to women with endometrioma and diffuse adenomyosis, which represented the large majority of cases (91.7% and 97.8%, respectively). Given their predominance, findings were largely consistent with those observed in the overall study population. In women with diffuse adenomyosis, higher progesterone levels were observed among those achieving live birth in HRT-FET cycles. In contrast, progesterone concentrations remained comparable between endometrioma patients who had a live birth and those who did not across both preparation protocols. Other disease subtypes were not analysed due to limited sample size. All clinical outcomes were analyzed using multivariable logistic regression models adjusting for age, body mass index, progesterone levels, estradiol levels, endometrial thickness, embryo quality, and preparation protocol (Table  5 ). Table 5 Clinical outcomes in women with endometriosis and adenomyosis compared with controls Controls (reference) Endometriosis Adenomyosis Rate (%) Rate (%) OR 95% CI p -value Rate (%) OR 95% CI p -value Biochemical pregnancy 43.32% 48.32% 0.96 0.72–1.28 0.767 46.24% 0.89 0.70–1.14 0.364 Implantation 41.00% 40.67% 0.95 0.74–1.22 0.698 41.50% 1.01 0.79–1.29 0.934 Clinical pregnancy 35.49% 35.78% 0.88 0.65–1.19 0.393 35.38% 0.95 0.74–1.22 0.687 Miscarriage 5.51% 4.89% 0.80 0.45–1.44 0.456 6.12% 1.07 0.63–1.81 0.805 Live birth 31.77% 32.11% 0.89 0.65–1.22 0.459 32.31% 0.99 0.76–1.27 0.912 Values are expressed as percentages. Odds ratios (OR) with 95% confidence intervals (CIs) and p -value were obtained from multivariable logistic regression models adjusted for age, AMH, body mass index, progesterone levels, estradiol levels, endometrial thickness, embryo quality (A as reference), and endometrial preparation protocol Clinical outcomes in women with endometriosis and adenomyosis compared with controls Values are expressed as percentages. Odds ratios (OR) with 95% confidence intervals (CIs) and p -value were obtained from multivariable logistic regression models adjusted for age, AMH, body mass index, progesterone levels, estradiol levels, endometrial thickness, embryo quality (A as reference), and endometrial preparation protocol No significant associations were observed between endometriosis or adenomyosis and biochemical pregnancy, implantation, clinical pregnancy, live birth, or miscarriage.

A single-center, retrospective observational cohort study was conducted at the Reproductive Medicine Center of IRCCS San Raffaele University Hospital (Milan, Italy) between June 2020 and December 2023. The study included women undergoing their first frozen embryo transfer (FET) cycle with single blastocyst transfer during the study period. Patients were subsequently classified according to the presence of endometriosis and/or adenomyosis based on imaging and/or surgical diagnosis. Unexposed controls were defined as women without ultrasound or clinical evidence of endometriosis or adenomyosis. Ultrasound evaluation was performed during pre-ART assessments by experienced sonologists according to standardized criteria proposed by the International Deep Endometriosis Analysis (IDEA) group [ 28 ]. Women with a previous diagnosis of endometriosis confirmed by transvaginal sonography (TVS), magnetic resonance imaging (MRI), or histological examination following surgery were also included. MRI diagnosis was based on established radiological features of endometriotic lesions, including endometriomas and deep infiltrating disease [ 29 ], while histological confirmation was defined by the presence of ectopic endometrial glands, stroma, and/or hemosiderin-laden macrophages in suspected lesions biopsied at the time of surgery [ 30 ]. Endometriotic lesions were classified into superficial peritoneal (SUP), ovarian endometrioma (OMA), or deep infiltrating endometriosis (DIE) phenotypes [ 31 ]. For cases of SUP, as the pre-ART evaluation did not reveal OMA or DIE lesions, the diagnosis relied on surgical findings. Adenomyosis was diagnosed using standardized TVS criteria based on morphological uterine features as defined by the Morphological Uterus Sonographic Assessment (MUSA) criteria [ 32 ], and, when available, supported by MRI findings based on established radiological features of adenomyosis [ 33 ]. Cases were subsequently classified into diffuse adenomyosis or adenomyoma according to established definitions [ 34 ]. All ultrasound examinations were performed by experienced operators. Disease severity was not systematically staged. In addition, diagnoses based on imaging and those confirmed surgically were not analyzed separately. Only the first FET of a single blastocyst per patient was analyzed. Embryo quality was assessed according to the Gardner and Schoolcraft grading system [ 35 ] and categorized into high-, intermediate-, and low-quality embryos for use in multivariable analyses. Cycles were excluded if the transfer was canceled, if parenteral progesterone was administered, if involved oocyte donation, if cleavage-stage embryos were transferred or if preimplantation genetic testing (PGT) was performed. Each patient therefore contributed one FET cycle, conducted either under HRT or mNC, chosen according to individual characteristics and clinical judgment rather than randomization. The primary outcome was serum progesterone level on the day of embryo transfer. Progesterone levels were compared across diagnostic groups (endometriosis, adenomyosis, and controls) and further evaluated according to endometrial preparation protocol and live birth status. Secondary outcomes included comparisons of reproductive outcomes (biochemical pregnancy, implantation, clinical pregnancy, miscarriage, and live birth) across the three diagnostic groups, as well as subgroup analyses of progesterone levels according to disease phenotype. Implantation was defined as the presence of a gestational sac on TVS, clinical pregnancy was defined as the presence of fetal cardiac activity on TVS, miscarriage was defined as loss of clinical pregnancy, including cases with a gestational sac without embryonic development and cases with absent fetal cardiac activity. HRT consisted of estrogen treatment beginning on Day 1 of menstruation, using transdermal estradiol (0.1 mg/day, simultaneously applying two Estraderm 50 patches; IBSA, Switzerland) or oral estradiol (6 mg/day, Progynova; Bayer, Germany). After 8–10 days of estrogen therapy, TVS was performed to measure endometrial thickness, and a blood sample was collected to assess serum progesterone levels, ensuring that no spontaneous ovulation had occurred. FET at the blastocyst stage was planned when endometrial thickness exceeded 7 mm and serum progesterone was < 1.5 ng/mL. Progesterone supplementation began five days before FET with daily vaginal progesterone at a dose of 800 mg (Amelgen 400 mg, 400 mg capsule twice daily; Gedeon, Italy). In mNC, spontaneous follicular growth was monitored by TVS on days 8–10 of the cycle to assess endometrial thickness. A blood sample was collected to measure serum progesterone levels and confirm the absence of premature ovulation. When endometrial thickness was ≥ 7 mm and a dominant follicle > 16 mm was visualized, ovulation was triggered by subcutaneous injection of 5000 IU hCG (human chorionic gonadotropin, Gonasi; IBSA, Italy). Progesterone supplementation began two days after hCG administration, with daily vaginal progesterone at a dose of 800 mg (Amelgen 400 mg, 400 mg capsule twice daily; Italfarmaco, Italy). In cases of serum progesterone levels below 10 ng/mL on the day of embryo transfer, rescue progesterone supplementation was administered [ 27 ]. On the day of FET, serum progesterone levels were measured between 10:00 and 11:00 a.m. Although blood sampling was performed within a fixed time window, the exact interval from the last vaginal progesterone administration could not be fully standardized. Blood samples were analyzed using an electrochemiluminescence immunoassay (Tosoh AIA fluorimetric system, ST-AIA-PACK, Tosoh Corporation, Tokyo, Japan) with a sensitivity of 0.1 ng/mL. The same assay platform was used consistently throughout the study period to ensure inter-assay reliability and minimize analytical variability. Baseline demographic and clinical characteristics were compared between groups. Patients were classified into three groups based on ultrasound findings: adenomyosis, endometriosis and controls. In cases of coexisting adenomyosis and endometriosis, patients were included in both diagnostic groups, as the analyses were conducted independently and no direct comparison was performed between them. In order to avoid potential bias related to non-mutually exclusive group allocation, a sensitivity analysis was performed using mutually exclusive categories: controls, endometriosis only, adenomyosis only, and combined endometriosis and adenomyosis. Within each diagnostic group, patients were further categorized according to the predominant phenotype: SUP, OMA or DIE for endometriosis, and adenomyoma or diffuse adenomyosis for adenomyosis. When multiple phenotypes coexisted within the same diagnostic group, patients were assigned to the most severe category [ 36 , 37 ]. These phenotype-based subgroupings were used exclusively for secondary analyses and did not modify the primary diagnostic group assignment. Given the known physiological differences in progesterone exposure between HRT and mNC, and the non-random allocation to endometrial preparation protocols, all analyses were stratified by protocol and no direct comparisons were performed between HRT and mNC cycles. All data were stored in a digital database and analyzed using RStudio (R version 4.2.2; R Foundation for Statistical Computing, Vienna, Austria). A p -value < 0.05 was considered statistically significant. Continuous variables were expressed as mean ± standard deviation (SD) for normally distributed data, or as median and interquartile range (IQR) for non-normally distributed data, while categorical variables as counts and percentages. Between-group comparisons were performed using univariable linear regression model for continuous variables and univariable logistic regression model for categorical ones. To account for the non-normal distribution of progesterone levels, a sensitivity analysis was conducted by repeating these comparisons using log-transformed progesterone values. Multivariable linear regression models were used to evaluate the association between patient characteristics and serum progesterone levels on the day of FET, while multivariable logistic regression models were used to assess the association between diagnostic groups and reproductive outcomes (biochemical pregnancy, implantation, clinical pregnancy, miscarriage, and live birth). Biochemical pregnancy was defined as a positive serum β-hCG test. Implantation was defined as the presence of a gestational sac on ultrasound. Clinical pregnancy was defined as the presence of an intrauterine gestational sac containing an embryo with fetal cardiac activity. Miscarriage was defined as pregnancy loss before 12 weeks of gestation and was calculated among patients with implantation. Live birth was defined as the delivery of a live neonate. Covariates included in the multivariable models were selected based on clinical relevance and prior evidence, rather than solely on univariable statistical significance, in order to avoid excluding potential confounders. In linear regression models, both univariable and multivariable, regression coefficients (β) represent the expected change in progesterone levels associated with a one-unit increase in the predictor variable. With a minimum of 237 participants per group, the study achieved 80% power to detect a 1 ng/mL difference in progesterone levels versus controls, assuming a two-sided α = 0.05. Subgroup analyses were underpowered and should be interpreted with caution.

In the present study, mean serum progesterone levels measured on the day of FET were comparable between women with endometriosis and/or adenomyosis and unaffected controls, irrespective of the endometrial preparation protocol. These findings indicate that, under standard luteal phase support, neither condition is associated with reduced systemic progesterone exposure at the time of embryo transfer. These findings are consistent with previous evidence from HRT-FET cycles. In a cohort exceeding 1,700 cycles, Bourdon and colleagues [ 26 ] reported comparable serum progesterone concentrations in women with and without endometriosis or adenomyosis, both among those achieving live birth and those not achieving live birth. Within this context, the present study expands the available evidence by including a large, well-characterized cohort and by systematically evaluating both HRT and modified natural cycles, thereby providing a more comprehensive and physiologically relevant assessment of progesterone exposure across different clinical settings. Taken together, these findings suggest that women with endometriosis or adenomyosis do not require universally higher circulating progesterone levels and support the concept that progesterone resistance at the tissue level does not necessarily translate into reduced systemic progesterone exposure during ART. These clinical observations should be interpreted in light of the extensive biological literature describing progesterone resistance at the tissue level. Endometriosis is characterized by impaired progesterone signaling, reduced expression of progesterone receptor isoform B, and altered downstream gene activation involved in decidualization [ 2 , 22 ]. Similar molecular alterations have been described in adenomyosis, where local estrogen excess, chronic inflammation, and dysregulated steroid receptor expression affect both ectopic and eutopic endometrium [ 2 , 3 ]. Our results suggest that these tissue-level abnormalities do not manifest as lower mean serum progesterone concentrations at the time of FET when conventional supplementation regimens are used. When focusing on adenomyosis, no overall difference in mean serum progesterone levels was observed compared with controls. However, a more nuanced pattern emerged within HRT-FET cycles: women with adenomyosis who did not achieve live birth exhibited lower serum progesterone levels compared with those who did, a trend that was also observed in diffuse disease. To our knowledge, this observation has not been consistently reported in previous studies and therefore adds a novel element to the current literature. Nevertheless, this finding should be interpreted cautiously and considered hypothesis-generating, given the observational design and the exploratory nature of subgroup analyses. From a clinical perspective, this may suggest that insufficient progesterone exposure could contribute to implantation failure in a subset of patients, potentially those with more severe forms of adenomyosis, rather than representing a uniform feature of the condition. Accordingly, individualized luteal phase support may be considered in selected cases, particularly in women with more extensive disease, although this hypothesis requires further prospective validation. Importantly, a recent study attempting to personalize luteal phase support in adenomyosis by adjusting progesterone timing based on endometrial receptivity analysis did not demonstrate improved reproductive outcomes [ 38 ], indicating that optimization of progesterone exposure alone is unlikely to overcome the complex endometrial dysfunction associated with this condition. Indeed, accumulating evidence suggests that adenomyosis impairs endometrial receptivity through multiple mechanisms, including chronic inflammation, altered gene expression, disruption of the junctional zone architecture, and abnormal uterine contractility, which are not fully corrected by increasing progesterone availability [ 39 – 41 ]. For patients with endometriosis, a recent large systematic review and meta-analysis demonstrated significantly reduced implantation and live birth rates in recipients with endometriosis, confirming the existence of a uterine factor even when oocyte contribution was eliminated [ 7 ]. Our data add an important dimension to this discussion suggesting that such a uterine factor is unlikely be mediated by reduced systemic progesterone exposure at the time of embryo transfer. Nevertheless, a recent study reports improved reproductive outcomes in women with endometriosis achieving higher serum progesterone levels during FET cycles [ 27 ], leaving open the possibility that progesterone requirements may be heterogeneous within this population. In our cohort, BMI emerged as a significant determinant of serum progesterone levels, in line with previous evidence linking higher body mass index to reduced progesterone absorption and lower circulating concentrations [ 42 ]. This observation was independent of disease status and highlights the importance of accounting for patient-specific characteristics when interpreting progesterone levels in clinical practice. Several limitations should be acknowledged. The retrospective, single-center design introduces the potential for selection bias and residual confounding. In addition, allocation to endometrial preparation protocols (HRT vs. mNC) was not randomized, and comparisons between protocols should therefore be interpreted as descriptive rather than causal. Patients with coexisting endometriosis and adenomyosis were initially included in both diagnostic groups, resulting in non-mutually exclusive categories; however, sensitivity analyses using mutually exclusive classifications yielded consistent results, supporting the robustness of the findings. Serum progesterone was assessed at a single clinically relevant time point on the day of embryo transfer; while reflective of routine practice, this approach does not capture cumulative luteal exposure, and variability related to the timing of the last administration could not be fully controlled. Moreover, circulating progesterone represents a surrogate of endometrial exposure, particularly in the context of vaginal administration and the “first-pass uterine effect”, whereby progesterone is preferentially delivered to the uterus, resulting in higher endometrial tissue concentrations compared with systemic levels. Consequently, serum progesterone may not fully reflect local endometrial exposure, and inter-individual variability in absorption and distribution cannot be excluded [ 43 , 44 ]. Although several relevant confounders were considered, residual confounding cannot be excluded, including factors related to embryo ploidy. In addition, subgroup analyses were relatively small and may have been underpowered to detect subtle differences. A subgroup analysis based on rescue progesterone supplementation was not performed, as this intervention may modify serum levels at the time of transfer and confound comparisons across groups. Finally, detailed information on prior treatments was not consistently available due to the retrospective design, potentially introducing some degree of heterogeneity within the study population. Prospective, adequately powered studies with standardized protocols, multiple progesterone measurements, and integrated assessment of systemic and endometrial exposure are warranted to better define the clinical relevance of progesterone thresholds and to evaluate whether individualized luteal phase support based on progesterone levels improves reproductive outcomes in patients with endometriosis or adenomyosis. Given the multifactorial nature of ART, integrating biological and psychosocial aspects into individualized patient management may further enhance treatment success and patient care [ 45 ]. In conclusion, our study provides evidence that women with endometriosis or adenomyosis do not exhibit lower systemic progesterone levels at the time of FET compared with unaffected controls. Despite well-documented progesterone resistance at the tissue level, conventional luteal phase support resulted in comparable serum progesterone concentrations and live birth rates across diagnostic groups. In adenomyosis, lower progesterone levels among women who did not achieve live birth suggest that a subset of patients may benefit from optimized luteal support, although this finding requires prospective validation. Overall, these findings support current ART practices while highlighting the need for further studies to better define individualized luteal phase support strategies in women with adenomyosis.

Endometriosis and adenomyosis are chronic gynecological disorders characterized by the presence of ectopic endometrial-like tissue outside and within the uterine wall, respectively, and frequently coexist in women of reproductive age. Their pathophysiology appears closely interconnected, involving estrogen-driven inflammation, altered uterine function, and impaired progesterone signaling, all of which may compromise endometrial receptivity despite exogenous hormonal support [ 1 – 3 ]. Despite this biological plausibility, the clinical impact of endometriosis and adenomyosis on assisted reproductive technology (ART) outcomes remains debated. In adenomyosis, systematic evidence suggests an association with poorer reproductive outcomes, particularly an increased risk of miscarriage, although findings vary according to disease phenotype and severity [ 4 , 5 ]. Studies in donor-oocyte recipients have reported preserved implantation rates but increased pregnancy loss, suggesting that uterine factors may play a relevant role [ 6 ]. In endometriosis, evidence derived from oocyte donation models and registry-based analyses indicates at most a modest reduction in live birth rates, supporting the hypothesis that impaired receptivity may be limited or heterogeneous across patient subgroups [ 7 ]. ART represents the main therapeutic option for infertility in women with endometriosis or adenomyosis. In this context, frozen-thawed embryo transfer (FET) has gained increasing relevance, as it allows embryo transfer in a more controlled endometrial environment and may reduce the adverse effects of controlled ovarian stimulation, such as supraphysiological estrogen exposure and inflammation, which may negatively affect implantation [ 8 – 12 ]. Current FET strategies include modified natural cycle (mNC) and hormone replacement therapy (HRT) protocols, with or without GnRH agonist suppression, although no approach has consistently demonstrated superiority in terms of live birth rates [ 13 , 14 ]. Regardless of the protocol adopted, adequate progesterone exposure remains a critical determinant of endometrial transformation and embryo–endometrium synchronization, and insufficient luteal phase support has been consistently associated with reduced implantation and live birth rates [ 15 – 21 ]. In this context, an additional concern arises for patients with endometriosis and/or adenomyosis, in whom progesterone resistance has been described at the tissue level [ 22 ]. In endometriosis, altered progesterone responsiveness has been reported primarily in the eutopic endometrium [ 23 ]. In adenomyosis, similar findings have been observed within adenomyotic lesions, while their presence in the eutopic endometrium remains debated [ 24 , 25 ]. Whether these observations are associated with different systemic progesterone requirements during frozen embryo transfer remains unclear. Bourdon and colleagues reported that patients with endometriosis and/or adenomyosis did not require higher progesterone levels on the day of HRT-FET to achieve live birth [ 26 ], whereas Alsbjerg and colleagues described higher serum progesterone levels in women with endometriosis undergoing HRT-FET [ 27 ]. Given these conflicting findings, and the increasing use of FET strategies in women with endometriosis and/or adenomyosis, further evidence is warranted. In particular, it remains unclear whether these conditions are associated with differences in systemic progesterone levels at the time of embryo transfer. Therefore, the present study aimed to evaluate serum progesterone levels on the day of embryo transfer in women with endometriosis and/or adenomyosis compared with unaffected controls. Analyses were conducted separately by endometrial preparation protocol, and associations with reproductive outcomes were assessed within each protocol, with additional subgroup analyses by disease phenotype.

Supplementary Material 1. Supplementary Table 1: Baseline demographic and clinical characteristics stratified by endometrial preparation protocol (HRT vs mNC). Supplementary Material 1. Supplementary Table 1: Baseline demographic and clinical characteristics stratified by endometrial preparation protocol (HRT vs mNC). Supplementary Material 2. Supplementary Table 2: Distribution of serum progesterone levels according to reproductive outcomes and disease status in HRT-FET and mNC-FET. Supplementary Material 2. Supplementary Table 2: Distribution of serum progesterone levels according to reproductive outcomes and disease status in HRT-FET and mNC-FET.