https://orcid.org/0000-0002-3048-9447
Abstract
Is there a relationship between serum and endometrial progesterone (P4) levels, including P4 and metabolites (oestrone, oestradiol and 17α-hydroxyprogesterone), and endometrial receptivity?
Serum P4 levels were not correlated with endometrial P4, nor associated with endometrial receptivity as determined by the ERA® test; however, endometrial P4 and 17α-hydroxyprogesterone levels were positively correlated and related to endometrial receptivity by ERA.
Acquisition of endometrial receptivity is governed by P4, which induces secretory transformation. A close relationship between serum P4 and pregnancy outcome is reported for hormone replacement therapy (HRT) cycles. However, the relationship between serum and uterine P4 levels has not been described, and it is unknown whether uterine receptivity depends more on serum or uterine P4 levels.
A prospective cohort study was performed during March 2018–2019 in 85 IVF patients undergoing an evaluation-only HRT cycle with oestradiol valerate (6 mg/day) and micronised vaginal progesterone (400 mg/12 h).
Patients were under 50 years of age, had undergone at least one failed IVF cycle, had no uterine pathology, and had adequate endometrial thickness (> 6.5 mm). The study was conducted at IVI Valencia and IVI Foundation. An endometrial biopsy and a blood sample were collected after 5 days of P4 vaginal treatment. Measures included serum P4 levels, ERA®-based evaluation of endometrial receptivity, and endometrial P4 levels along with metabolites (oestrone, oestradiol and 17α-hydroxyprogesterone) measured by ultra-performance liquid chromatography–tandem mass spectrometry.
Seventy-nine women were included (mean age: 39.9 ± 4.6, BMI: 24.2 ± 3.9 kg/m2, endometrial thickness: 8.2 ± 1.4 mm). The percentage of endometria indicated as receptive by ERA® was 40.5%. When comparing receptive versus non-receptive groups, no differences were observed in baseline characteristics nor in steroid hormones levels in serum or endometrium. No association between serum P4 and endometrial steroid levels or ERA result was found (P < 0.05). When the population was stratified according to metabolite concentration levels, endometrial P4 and 17α-hydroxyprogesterone were significantly associated with endometrial receptivity (P < 0.05). A higher proportion of receptive endometria by ERA was observed when endometrial P4 levels were higher than 40.07 µg/ml (relative maximum) and a lower proportion of receptive endometria was associated with endometrial 17α-hydroxyprogesterone lower than 0.35 ng/ml (first quartile). A positive correlation R2 = 0.67, P < 0.001 was observed between endometrial P4 and 17α-hydroxyprogesterone levels.
This study did not analyse pregnancy outcomes. Further, the findings can only be extrapolated to HRT cycles with micronised vaginal progesterone for luteal phase support.
Our findings suggest that the combined benefits of different routes of progesterone administration for luteal phase support could be leveraged to ensure an adequate concentration of progesterone both in the uterus and in the bloodstream. Further studies will confirm whether this method can optimise both endometrial receptivity and live birth rate. Additionally, targeted treatment to increase P4 endometrial levels may normalise the timing of the window of implantation without needing to modify the progesterone administration day.
This research was supported by the IVI-RMA Valencia (1706-VLC-051-EL) and Consellería d’Educació, Investigació, Cultura, i esport Generalitat Valenciana (Valencian Government, Spain, GV/2018//151). Almudena Devesa-Peiro (FPU/15/01398) and Cristina Rodriguez-Varela (FPU18/01657) were supported by the FPU program fellowship from the Ministry of Science, Innovation and Universities (Spanish Government). P.D.-G. is co-inventor on the ERA patent, with non-economic benefits. The other authors have no competing interests.
NCT03456375.
Introduction
Molecular synchronisation between the embryo and the endometrium is essential for successful implantation. Although embryo quality is critical in this process, the endometrium is also an important factor for embryo implantation (Somigliana et al., 2018; Pirtea et al., 2021). The endometrium must achieve a state of receptivity to the embryo. This process is governed by progesterone, which increases gradually during the luteal phase and induces the secretory transformation of the endometrium.
Exogenous progesterone is often administered as part of assisted reproductive techniques to compensate for a luteal phase defect. In artificial cycles, also called HRT cycles, there is no endogenous progesterone production due to the absence of corpus luteum. A strong relationship between P4 concentration and pregnancy outcome was recently demonstrated when micronised vaginal progesterone (MVP) was used for luteal phase support in HRT cycles (Labarta et al., 2017) and patients with serum P4 levels below 8.8 ng/ml had 18% lower chances of live birth than patients with serum P4 levels above this cut-off (Labarta et al., 2021). MVP administration induces a rapid increase in uterine P4 due to the first uterine pass effect (Bulletti et al., 1997; De Ziegler et al., 1997), which suggests rapid absorption of the hormone in the uterus after vaginal P4 administration. However, some studies suggest that the endometrium is dependent on arterial blood carrying P4 (intramuscular, subcutaneous, corpus luteum, vaginal or rectal pessary), and the endometrium receives P4 from vaginal pessaries as a standard mechanism (Yovich et al., 2015). Serum P4 reflects the concentration of progesterone within the vasculature. However, it is undetermined whether systemic or endometrial levels of progesterone are more important in determining endometrial receptivity.
Several markers identify endometrial receptivity, using techniques that are either conventional (e.g. ultrasound or histology) or newly developed (e.g. gene expression profiling) (Altmäe et al., 2014; Sebastian-Leon et al., 2018; Craciunas et al., 2019; Hernández-Vargas et al., 2020). One recent development is the ERA® test, which combines transcriptomics from an endometrial biopsy with artificial intelligence algorithms to objectively date the endometrium in the menstrual cycle phase (Díaz-Gimeno et al., 2011, 2013, 2017). This approach has been used in clinical practice for individualising the embryo transfer day to match endometrial receptivity, also known as the personalised window of implantation (WOI) (Ruiz-Alonso et al., 2013, 2014). Indeed, WOI timing can be modulated, using ERA® test results, by adjusting the timing of P4 administration in the next artificial cycle (later in pre-receptive and earlier in post-receptive endometria) (Ruiz-Alonso et al., 2013, 2014). This technology therefore offers the potential to uncover the relative contributions of serum versus uterine P4 to endometrial receptivity. To capture this opportunity, we measured serum and endometrial levels of P4 and its metabolites on the day of endometrial biopsy in a cohort of women undergoing HRT and evaluated the relationship between progesterone measures and ERA test results.
Materials and methods
Ethics statement
This study was performed at the Instituto Valenciano de Infertilidad (IVI), Valencia, Spain during March 2018–2019. All eligible patients were offered participation in the study and those interested provided written informed consent. The study was approved by the Institutional Review Board at Instituto Valenciano de Infertilidad (1706-VLC-051-EL); www.clinicaltrials.gov registration number NCT03456375.
Study design and participants
This was a prospective cohort study comparing metabolite levels in infertile patients with receptive and non-receptive ERA® test results under an evaluation-only artificial endometrial preparation cycle with oestradiol valerate and micronised vaginal progesterone.
Patients were <50 years old with a history of one or more failed IVF cycles after transfer of good quality embryos and endometrial thickness >6.5 mm at the end of the proliferative phase, all of them with the same regular doses of oestradiol valerate. ERA® test was performed during the mid-secretory phase. Patients with any type of uterine pathology (fibroids, polyps, or congenital Mullerian abnormalities) or hydrosalpinx were excluded.
Endometrial preparation and sample collection
Oestradiol valerate was administered orally from the first or the second day of menstruation (6 mg/day of Meriestra®, Novartis, Barcelona, Spain). After 10 days of oestrogen treatment, a 2-dimensional vaginal ultrasound was performed to measure endometrial thickness and to confirm the three-layer endometrial pattern. A blood sample was drawn to measure oestradiol (E2) and P4 to ensure that no spontaneous ovulation had occurred. If endometrial thickness was ≥6.5 mm, the endometrial pattern was trilaminar, and serum P4 was less than 1.0 ng/ml, the endometrial biopsy was scheduled (Fig. 1A). Luteal phase support began 5 days before the biopsy with MVP at a dose of 400 mg every 12 h (Utrogestan, SEID, Barcelona, Spain or Progeffik, Effik, Madrid, Spain). Endometrial biopsy was done after 5 days of vaginal progesterone, based on the expected WOI timing in a standard cycle (Fig. 1A).
Study design and patient population. (A) Experimental design. Upper part of the figure shows an endometrium and represents endometrial progression across the menstrual cycle days (Day 0 (first day of menstruation). Black and grey bars indicate doses of vaginal progesterone and oestradiol valerate administrated to the patient in a hormone replacement therapy cycle. Endometrial biopsies were collected in the expected window of implantation (WOI) 5 days after progesterone administration (P + 5). Ultrasound was performed 10 days after the first day of menstruation and only patients with normal endometrium (trilaminar pattern, endometrial thickness higher than 6.5 mm with the regular doses of oestradiol and serum P4 lower than 1 ng/ml) were biopsied in P + 5 (Day 16) and included in the study. This biopsy was used to perform the ERA® test and the remaining part was used for measuring endometrial steroid hormones metabolites (P4, 17α-hydroxyprogesterone, E2 and oestrone). E2: oestradiol or 17β-oestradiol; P4: progesterone. (B) Demographics and population homogeneity related to the study variables. Assuming a normal distribution for all variables, a t-test was implemented to evaluate the differences in the receptivity groups. R, receptive patients according to ERA® test; NR, non-receptive patients according to ERA® test; P, P value; N, number of patients. In the variable column, are the units and the menstrual cycle day when the variable was measured: 10th is 10 days after the first day of menstruation. P + 5: 5 days after progesterone administration in hormone replacement therapy. 17α-OHP, 17α-hydroxyprogesterone.
All patients underwent endometrial biopsy of the uterine fundus collected under sterile conditions using a Cornier pipelle cannula (CCD Laboratories, Paris, France). First, progesterone residues were removed from the cervix. Once introduced through the cervical canal, the biopsy was obtained by scraping down the endometrium from the bottom of the uterus and rotating the catheter so that tissue was taken from all the walls of the endometrial cavity. After biopsy, endometrial tissue was sent to IVI Foundation for measurement of the following steroid hormones: P4, 17α-hydroxyprogesterone, E2, and oestrone. A blood test was performed at the time of the endometrial biopsy (±1 h) to determine serum P4 levels, 6–8 h after the last dose of MVP.
Serum and endometrial steroid hormone measurement
Blood samples were analysed by an electrochemiluminescence immunoassay (Cobas e411 analyser, Roche diagnostics GmbH, Germany). Intra- and inter-assay coefficients of variation for the P4 determinations were 1.2–11.8% and 3.6–23.1%, respectively, for P4 values between 0.22 and 51.6 ng/ml. Sensitivity was 0.03 ng/ml. The intra- and inter-assay coefficients of variation for E2 determinations were 2.4–9.5% and 2.5–11.9%, with a measurement range of 25.4–2932 pg/ml. The sensitivity was 5 pg/ml.
Endometrial P4, 17α-hydroxyprogesterone, E2 and oestrone were measured using ultra performance liquid chromatography–tandem mass spectrometry (UPLC-MS/MS), with steroid metabolites selected according to their detection in human endometrium (Häkkinen et al., 2018). Approximately 50–100 mg of each frozen endometrial biopsy were placed in 2 ml tubes containing CK14 ceramic beads (Precellys). For each 50 mg of tissue, 300 μl of ethyl acetate were added and tissues were homogenised twice for 40 s at 6000 rpm at 4°C in a Precellys 24 Dual system equipped with a Criolys cooler (Precellys). Tubes were then centrifuged at 3000g for 5 min at 4°C, and supernatants were transferred to clean tubes. A second extraction was performed with 200 μl of ethyl acetate/50 mg of tissue and evaporated to dryness in a Savant speedvac concentrator. Samples were reconstituted in 65 μl of water: acetonitrile (70:30, v/v), centrifuged at 10 000g for 10 min at 4°C, and supernatants were filtered and transferred into 96-well plates for further analysis. Analysis was performed using two internal standards, progesterone-d9 and 17β-oestradiol-2,3,4-13C3 (Sigma-Aldrich, Merk, Germany), on an Acquity UPLC system (Waters, Wilmslow, UK) equipped with an Acquity UPLC BEH C18 column (1.7 μm, 2.1 × 100 mm; Waters) with water and acetonitrile with 0.1% formic acid as mobile phases. Mass spectrometry analysis was performed using a Waters Xevo TQ-S mass spectrometer (Waters) with an electrospray ionisation source working in positive-ion mode and multiple reaction monitoring mode.
Receptivity measurement
The current ERA® test is based on an RNA-Seq panel that evaluates the expression of 236 genes (igenomix webpage) and classifies endometrial samples considering seven possible profiles according to endometrial timing as pre-receptive 48 h (PRE48), pre-receptive 24 h (PRE24), early-receptive 12 h (ER12), receptive on time (ROT), late-receptive -12 h (LR12), post-receptive -24 h (POST-24) and post-receptive -48 h (POST-48). Biopsies were classified according to these criteria with ROT considered receptive and the remaining categories (PRE48, PRE24, ER12, LR12, POST-24, POST-48) considered non-receptive.
Statistical analysis
Normal distributions were assumed for demographics, clinical variables and steroid levels due to the sample size being higher than 50 in all the statistical contrasts. All variable effects related to receptivity groups were independently compared using a Student’s t-test. The primary endpoint was the relationship between serum and tissue P4, estrone, oestradiol and 17α-hydroxyprogesterone levels and ERA® diagnoses (receptive vs non-receptive). Secondary endpoints were the relationship between serum and tissue metabolites and the relationships among metabolites.
After an exploratory analysis of concentration distribution in the population for each metabolite, serum P4 and endometrial levels of steroid hormones (P4, 17α-hydroxyprogesterone, E2, estrone) were tested for association with receptivity by ERA®.
Hormone thresholds were calculated using first and third quartile (Q1 and Q3) as well as relative maximum (defined as the third quartile plus 1.5 times the interquartile distance). Patients were then stratified into two groups to compare patients over or below each threshold. Additionally, we used the threshold of 8.8 ng/ml for serum P4 based on our previous clinical data (Labarta et al., 2021) for correlation with ERA® diagnosis. The relationship between each group and ERA® diagnosis (primary endpoint) was determined using a non-parametric Barnard’s test (Erguler, 2016) to avoid loss of power due to greater discreteness of Fisher’s statistic. Proportion differences with P values less than 0.05 were considered statistically significant. Relationships between serum and tissue metabolites and between tissue metabolites (secondary endpoints) were evaluated using Pearson’s correlation test. All statistical analyses and file processing were implemented in R statistical software version 3.3.2 (R Core Team, 2016).
Results
We recruited 85 patients, of whom 79 underwent a successful endometrial biopsy and molecular analysis. Reasons for cancellation were an endometrial sample of inadequate quality (n = 3), no blood extraction on the day of biopsy (n = 1) and self-withdrawal for personal reasons during endometrial preparation (n = 2). The study population had a mean age of 39.9 ± 4.6 years, a body mass index (BMI) of 24.2 ± 3.9 kg/m2, and an endometrial thickness of 8.2 ± 1.4 mm during the proliferative phase (Fig. 1B).
ERA® identified a receptive profile in 40.5% of patients (n = 32); the remainder were non-receptive (n = 47). Among the non-receptive cases, 27.7% were early-receptive (PRE48), 65.9% were pre-receptive (PRE24), 4.3% were late-receptive (LR12) and 2.1% were post-receptive (POST24).
The study groups (receptive versus non-receptive) were homogeneous and comparable in age, BMI, endometrial thickness and proliferative-phase hormone levels (P > 0.05). Mean levels of steroid hormones in serum and endometrial tissue on the day of endometrial biopsy did not differ between patients diagnosed as receptive or non-receptive (Fig. 1B). We checked for a non-age effect on metabolites concentrations (data not shown).
Mean serum P4 levels on the day of endometrial biopsy were 13.52 ± 5.5 ng/ml. Stratification according to serum P levels showed a lack of association with endometrial receptivity according to ERA® (Fig. 2). The percentages of receptive by ERA® were 37.5% (6/16) and 41.3% (26/63) for serum P4 below or above 8.8 ng/ml on the day of endometrial biopsy, respectively (P = 0.810) (Fig. 2).
Serum progesterone levels are not related to ERA® test timing. (A) Patient stratification according to metabolite levels in the expected WOI (P + 5). Histogram for metabolite concentrations in our patient population. Vertical dotted lines represent the different thresholds used to divide patients in two groups. Purple represents the additional threshold for serum P4 based on our previous study based on clinical data (Labarta et al., 2021). Orange represents the first quartile (Q1), blue represents the third quartile (Q3), and green indicates the relative maximum calculated from distribution of serum progesterone concentration measured for 79 patients. Data are divided in two groups according to each of these thresholds. Arrows below this histogram represent the different partitions of data into two groups (one group are patients with serum P4 concentration above the threshold and the other group are patients with serum P4 concentration below of the threshold), using the different thresholds. Number above each arrow indicates the number of patients in each group according to the considered threshold. (B) Bar plots representing the percentage of receptive (green) and non-receptive (orange) patients according to the ERA® test on each group (above or below the threshold). The threshold represented on each bar plot is indicated in the title. Inside each bar is indicated the number of patients being receptive (green bars) or non-receptive (orange bar) as well as the percentage of the group that represent this number of patients (into brackets). Barnard’s test was applied for contrasting the statistical significance for the group’s differences in receptivity distribution. P is shown for each comparison. Significant P values are indicated by *.
A higher proportion of receptive ERA® profiles was observed in patients with endometrial P4 levels higher than 40.07 µg/ml (77.8% vs 34.8% for those with levels lower than this threshold, which corresponded to the maximum relative value) (P = 0.046) (Fig. 3A). Similarly, a lower proportion of receptive ERA® profiles was observed when endometrial 17α-hydroxyprogesterone was lower than 0.35 ng/ml (12.5% vs 44.9%), which corresponded to the first quartile (P = 0.026) (Fig. 3B). Endometrial oestrogen-related metabolites (oestrone and oestradiol) were not associated with endometrial receptivity (Supplementary Fig. S1).
Endometrial progesterone-related metabolites are related to ERA® test timing. (A) Tissue progesterone. (B) Tissue 17α-hydroxyprogesterone. The upper part of the figure represents a histogram for metabolite concentrations in our patient population. Vertical dotted lines indicate the thresholds (Q1, Q3 and relative maximum) calculated from steroid hormone concentration data (tissue progesterone or 17α-hydroxyprogesterone). Orange represents the first quartile (Q1), blue represents the third quartile (Q3), and green indicates the relative maximum. Arrows below histogram represent the three partitions of the patients according to the different thresholds defined into two groups (above or below of the considered threshold). Colours are the same as the thresholds indicated as vertical dotted lines in the histogram. Numbers on each arrow indicate the number of patients in each group above and below the corresponding threshold. Bars plots in the lower part of the figure represent the percentage of receptive (green) and non-receptive (orange) patients according to the ERA® test on each group (above or below the threshold). The threshold represented on each bar plot is indicated in the title. Inside each bar is indicated the number of patients being receptive (green bars) or non-receptive (orange bar) as well as the percentage of the group that represents this number of patients (into brackets). Barnard’s test was applied for contrasting the statistical differences in receptivity proportion between the two groups. P value is shown for each comparison. Significant P values are indicated by *.
No correlation was observed between serum P4 levels at the time of endometrial biopsy and any of the endometrial metabolites. However, endometrial 17α-hydroxyprogesterone levels were correlated with both endometrial P4 (R2 = 0.67, P < 0.001) and estrone (R2 = 0.27, P = 0.048) (Fig. 4A). Because endometrial P4 and 17α-hydroxyprogesterone were correlated, there were no patients with high P4 levels (>40.07 µg/ml) and low 17α-hydroxyprogesterone (<0.35 ng/ml) (Fig. 4B). Indeed, five of the seven (71% of) patients with endometrial P4 levels higher than 40.07 µg/ml and with 17α-hydroxyprogesterone higher than 0.35 ng/ml were diagnosed as receptive, while only two of 15 (13% of) patients with P4 levels less than 40.07 µg/ml and with 17α-hydroxyprogesterone less than 0.35 ng/ml were diagnosed as receptive (Fig. 4B).
Steroid-related metabolite correlations and progesterone metabolite behaviour. (A) Pearson correlations between serum P4 and endometrial tissue steroid related metabolites and between endometrial metabolites. Significant correlations are indicated from one to three asterisks. (B) Relationship between progesterone and 17α-hydroxyprogesterone (17α-OHP) in the patient population.
Discussion
This is the first study to analyse the relationship between serum and endometrial progesterone levels and endometrial receptivity according to the ERA® test. Importantly, serum P4 levels were not correlated with endometrial P4 in our samples, nor were they associated with ERA® results. However, uterine P4 and 17α-hydroxyprogesterone showed a positive correlation and were statistically related with endometrial receptivity by ERA®. Notably, while the potential utility of performing personalised embryo transfer based on ERA® results (Patel et al., 2019; Simón et al., 2020) remains controversial (Bassil et al., 2018; Cozzolino et al., 2020; Hirschberg et al., 2021; Lensen et al., 2021), our study leveraged the accuracy of ERA gene expression as an objective measure of the WOI to uncover the contributions of serum versus uterine P4 to receptivity.
Because serum P4 in the mid-luteal phase influences pregnancy outcome (Labarta et al., 2017), we sought to determine whether the impact of serum P4 on pregnancy is due to a direct effect on the endometrium during the luteal phase (enhancing embryo implantation) or due to systemic effects related to immunomodulation in later pregnancy. In our study, there was no relationship between serum P4 and endometrial receptivity, but we cannot ensure the absence of a causal link due to the statistical power of the study and the method we used to determine receptivity. However, in this same context, endometrial P4 was significantly more related to receptivity than was serum P4.
Interestingly, although serum P4 levels were not related to endometrial metabolites or to the ERA® profile, our recent clinical study confirmed a minimum level of serum P4 needed to maintain ongoing and live birth rates in HRT cycles for embryo transfer (Labarta et al., 2021). In fact, patients with serum P4 < 8.8 ng/ml had significantly lower live birth rates and higher clinical miscarriage rates. The proportion of non-receptive by ERA® was similar in patients with serum P4 below or above 8.8 ng/ml, suggesting that this amount of systemic P4 is crucial to mediate the immunological adaptations needed to maintain gestation (Shah et al., 2019), but not to acquire receptive status, since receptivity can be achieved with low serum P4 levels (Young et al., 2017).
On the other hand, our current study suggests that endometrial P4, but not systemic P4, could influence the results of the ERA® test. Indeed, patients with high levels of P4 and 17α-hydroxyprogesterone showed a higher proportion of receptive by ERA. This implies that if sufficient uterine levels of progesterone are reached regardless of serum levels, receptivity is achieved, although some clinical data do not support these findings (Mohammed et al., 2019). Previous studies demonstrated that uterine P4 levels were significantly higher when P4 was administered vaginally compared to intramuscularly (Paulson et al., 2014), explained by the first uterine pass effect (Bulletti et al., 1997; De Ziegler et al., 1997). Conversely, serum P4 levels are significantly higher when P4 is administered intramuscularly compared to vaginally. However, it is difficult to interpret whether a high level of P4 in the uterus or in the bloodstream is most physiologically applicable. Recent reviews suggest better results with intramuscular injection, and this could reinforce the relevance of having high serum P4 levels (Devine et al., 2018; Mohammed et al., 2019). Potentially, a combined administration of both intramuscular and vaginal P4 could harness the benefits of both methods by producing high serum P4 levels (favouring pregnancy maintenance) and high endometrial P4 levels (favouring endometrial receptivity acquisition); however, this strategy has yet to be evaluated.
We found very high levels of tissue progesterone in comparison to Poutanen's team study that measured progesterone in endometrium with the same technology as us (by LC-MS/MS) (Huhtinen et al., 2014). One hypothesis for why these higher levels were detected could be that our patients were receiving vaginal progesterone administration in a hormone replacement therapy cycle. The vaginal administration produces a rapid increase in uterine levels due to the first uterine pass effect as explained above. The exogenous progesterone used was micronised, and this increases and favours the vaginal absorption of progesterone. This is not observed when using parenteral progesterone, when uterine levels of progesterone are significantly lower.
Progesterone induces the secretory transformation of the endometrium, but endometrial development proceeds normally even with low mid-luteal serum progesterone (Young, 2013). Low serum P4 can make the endometrium receptive but may not enough to favour placental formation and maintain pregnancy. The criteria used to evaluate pre-decidual changes based solely on histology have been criticised (Coutifaris et al., 2004; Murray et al., 2004; Díaz-Gimeno et al., 2013), although an experimental model examining the effects of varying serum P concentrations on endometrial histologic development and gene expression in ovulatory women (Young et al., 2017) has demonstrated that endometrial histologic dating does not accurately reflect their differences in circulating P4 and consequently is not a useful measure of luteal function. However, the study also observed differences in patterns of endometrial gene expression which are dependent on concentrations of serum P4. We did not observe this dependency in our study, perhaps because endometrial receptivity in the ERA® test is a dichotomous outcome (receptive or non-receptive by ERA) rather than multiple comparisons of the whole gene expression profile that could lead to varying levels of endometrial receptivity within the WOI (Craciunas et al., 2019). The only strategy proposed to promote endometrial receptivity is to use the ERA® result to modify the progesterone delivery regimen. Based on our results, we speculate that targeted treatment to increase endometrial P4 levels could normalise the timing of the WOI without needing to modify the progesterone administration days.
We found no correlation between serum and endometrial P4 levels. This lack of correlation between serum and uterine levels is in accordance with previous studies (Guerrero et al., 1975; Miles et al., 1994) and could be explained by intracrinology, based on the local regulation of steroid hormones in the endometrium through molecular mechanisms independent of regulating systemic steroid hormone serum levels (Gibson et al., 2018). The concept of intracrinology suggests that a proportion of steroid hormones are synthesised locally in peripheral target tissues from the precursors. In women without ovarian activity, active sex steroids are made in target tissues by an intracrine mechanism. Our patient population was under an artificial endometrial preparation cycle, with inhibition of ovulation, and thus, only exposed to the exogenous administration of steroid hormones. Taking into consideration that all patients received the same doses of exogenous P and that serum P levels were not related with the receptive status of the endometrium (according to ERA test), our findings suggest that progesterone and its metabolites can be metabolised in a variable manner in different patients as the intracrine mechanisms can vary among them.
When we analysed hormone levels for all metabolites (in serum and endometrium) according to endometrial timing, no significant differences were found. However, when we stratified patients according to quantile ranks of steroid hormone levels, we identified a relationship between progesterone endometrial metabolites and ERA® results. This stratification led to small groups for comparison, and future studies should include a greater number of patients to better analyse these findings. Furthermore, we cannot extrapolate these findings to the spontaneous ovulatory cycle or other types of endometrial preparation. This study also was not designed to test any association with pregnancy outcome.
Finally, endometrial receptivity is a complex and multifactorial process mediated by a multitude of molecules, many of which are regulated cyclically by menstrual cycle hormones (Craciunas et al., 2019). In addition to endometrial hormone metabolites, coactivators and inflammatory mediators, among others, participate in the transduction of the progesterone-induced response (Marquardt et al., 2019), which can be impaired by factors related to progesterone resistance, i.e. a decreased responsiveness of the endometrium to available P4 (Chrousos et al., 1986). There may also be more mechanisms implicated in the establishment of endometrial receptivity that may have influenced our results.
In this study, our aim was to measure serum P4 following the same molecular technique as it is measured in our daily clinical practice since we have previously found a significant clinical impact on pregnancy outcome when serum P levels were below a certain threshold measured by electrochemiluminescence immunoassay (Labarta et al., 2017, 2021).
However, in this study, we were comparing serum progesterone levels analysed by electrochemiluminescence immunoassay to progesterone levels measured by LC-MS/MS for the endometrium. The comparison between progesterone from both sources, and the comparison of metabolites concentrations with receptivity did not affect the study conclusions and findings.
In conclusion, we determined that serum P4 levels were not correlated with endometrial P4 in a study of women undergoing HRT, nor were P4 serum levels associated with ERA® results. These findings suggest that the role of serum P4 on clinical outcome may depend on additional systemic signals. We are the first, however, to identify a positive association between uterine concentrations of P4 and 17α-hydroxyprogesterone, and with a receptive ERA® result, suggesting that uterine P4 levels can influence receptivity status. Further studies are needed to demonstrate whether targeted treatment to increase endometrial P4 levels can normalise the WOI timing without needing to modify the progesterone administration days.
Supplementary data
Supplementary data are available at Human Reproduction online.
Data availability
The data underlying this article are available in the article and in its online supplementary material.
Acknowledgements
The authors thank IVI-RMA IVI Valencia and the IVI Foundation, Instituto de Salud Carlos III Ministry of Health, Ministry of Science, Innovation and Universities and Generalitat Valenciana for research support. Furthermore, the authors thank the study participants and the clinical staff who assisted with patient recruitment and organisation. Thanks also goes to Unidad Analítica del IIS La Fe, especially to Dr Marina Lopez Nogueroles for her technical support in metabolites measurement by LC-MS/MS, and Laura Caracena, study coordinator at IVI RMA Valencia for the unconditional support.
Authors’ roles
Study design and patient population were defined by E.L, P.C., E.B and P.D.-G. Patient recruitment was coordinated by E.L. with the help of C.V., J.G., P.C. and E.B. Sample collection and pre-processing were carried out by E.L. with the help of C.R.-V. and A.D.-P. supervised by P.D.-G. Metabolite measurement was coordinated by E.L. and P.D.-G. with the help of A.D.-P. and C.R.-V. Data analysis was done by P.S.-L. with the help of A.D.-P. and P.C. supervised by P.D.-G. and E.L. Clinical interpretation and conclusions were provided by E.L. and P.D.-G. with the help of E.B. Tables and figures were implemented by P.S.-L. and P.D.-G., supervised by E.L. The manuscript was written by E.L. and P.D.-G. with the help of P.S.-L. and A.D.-P. and revised by all authors (P.C, C.R.-V., C.V., J.G. and E.B.).
Funding
This research was supported by the IVI-RMA Valencia (1706-VLC-051-EL) and Consellería d’Educació, Investigació, Cultura, i esport Generalitat Valenciana (Valencian Government, Spain, GV/2018//151). Almudena Devesa-Peiro (FPU/15/01398) and Cristina Rodriguez-Varela (FPU18/01657) were supported by the FPU program fellowship from the Ministry of Science, Innovation and Universities (Spanish Government).
Conflict of interest
P.D.-G. is co-inventor on the ERA patent, with non-economic benefits. The other authors have no conflict of interests.
