{"paper_id":"f4761b8e-d116-411a-ba9b-5f704d33708f","body_text":"Despite its many advances and achievements, reproductive medicine has long neglected the\nendometrial factor. Indeed, since the inception of this field, the oocyte/embryo has\nremained the central focus. In contrast, the maternal endometrium was considered a passive\npart of the reproductive process: a ‘good embryo’ (or four or five) was all that mattered.\nYet, while embryology and embryo transfer technologies have improved considerably over the\npast 30 years, the efficacy of IVF remains low worldwide, with current live birth rates of\n25–30% per started cycle ( Adamson  et\nal. , 2018 ). At least part of this gap may derive from a failure to\nconsider the endometrium; after all, it is fair to say that any process relying on a\ncollaboration between partners requires the function and coordination of both.\nFurther progress in reproductive medicine, like in all of medicine, depends on bringing new\ntechnologies and concepts to bear on long-standing problems. In recent decades,\ntranscriptomics or RNA sequencing, has emerged as a powerful tool for clinical diagnosis of\ndisease ( Byron  et al. ,\n2016 ). Applications of transcriptomics are found in cancer ( Ferreira  et al. , 2014 ;  Tan  et al. , 2016 ), cardiovascular pathologies\n( Matsa  et al. , 2016 ) and\nneurodegenerative diseases ( Ferreiro  et\nal. , 2012 ), among others. The reproductive medicine field is no\nexception.\nThe endometrial receptivity analysis (ERA) was first published ten years ago ( Díaz-Gimeno  et al. , 2011 ) after\nmore than ten years of basic and translational research by a handful of pioneers, including\nour group. The research objective was to consider the endometrial factor and determine the\npotential to personalise this in the IVF workup, to ultimately synchronise embryo transfer\nto a receptive maternal endometrium. Since then, personalised medicine for the endometrial\nfactor has taken off, changing the clinical practice of more than 4000 reproductive clinics\nin more than 90 countries worldwide. Below, we summarise the concepts, data and clinical\napplications for the ERA.\n\nIn the 2000s, endometrial dating by histological evaluation ( Noyes  et al. , 1950 ) was used as a predictor of\nendometrial receptivity or fertility status ( Coutifaris  et al. , 2004 ;  Murray  et al. , 2004 ). This led to an absence of any reliable\ndiagnostic test to determine the endometrial status. Consequently, the standard workup for\ninfertility in clinics worldwide no longer included endometrial status, beyond a limited use\nof imaging to determine endometrial thickness and pattern. The frequently reported cut-off\nof 7 mm seems not to be justified to decide on cycle cancellation or to refrain from further\nIVF, nor to guide embryo transfer ( Kasius  et\nal. , 2014 ).\nWith the arrival of the genomics revolution, endometrial biology became deeply scrutinised.\nFour independent groups simultaneously reported on transcriptomic profiling of the secretory\nphase of the human endometrium in natural cycles, searching for the window of implantation\n(WOI) ( Kao  et al. , 2002 ;\n Carson  et al. , 2002 ;  Riesewijk  et al. , 2003 ;  Mirkin  et al. , 2005 ). Two other\ngroups extended this transcriptomic characterisation across the menstrual cycle ( Borthwick  et al. , 2003 ;  Ponnampalam  et al. , 2004 ).\nSubsequent studies were extended to ovarian stimulation cycles ( Mirkin  et al. , 2004 ;  Horcajadas  et al. , 2005 ;  Simon  et al. , 2005 ), and even refractory cycles\nin patients with inert intrauterine devices (IUD) ( Horcajadas  et al. , 2006 ) (for review see  Horcajadas  et al. , 2007 ). Since 2005, myriad\npapers have further described the transcriptomic profile across the menstrual cycle ( Mirkin  et al. , 2005 ;  Punyadeera  et al. , 2005 ;  Simon  et al. , 2005 ;  Yanaihara  et al. , 2005 ;  Talbi  et al. , 2006 ;  Critchley  et al. , 2006 ;  Horcajadas  et al. , 2008 ;  Haouzi  et al. , 2009 ;  Kuokkanen  et al. , 2010 ;  Tseng  et al. , 2010 ;  Van Vaerenbergh  et al. , 2010 ;\n Revel  et al. , 2011 ). The\nnext step was a comparison of endometrial profiles between fertile patients and those with\npathologies such as recurrent implantation failure ( Tapia  et al. , 2008 ;  Koler\n et al. , 2009 ;  Altmäe\n et al. , 2010 ;  Macklon,\n2017 ), endometrial cancer ( Habermann\n et al. , 2011 ), endometriosis ( Matsuzaki, 2011 ;  Garcia-Velasco, 2015 ), and obesity ( Comstock  et al. , 2017 ). This progress thereby facilitated the\ntransition from anatomical to molecular medicine of the endometrial factor and ultimately\npaved the way for its clinical application.\n\nThe ERA was the first transcriptomic test developed to diagnose the endometrial receptivity\nstatus of infertile patients ( Díaz-Gimeno  et\nal. , 2011 ). To identify genes involved in the human endometrial\nreceptivity signature, we initially analysed differences in genome-wide expression profiles\nbetween receptive and pre-receptive endometrium using raw expression data from three\ndifferent models of endometrial receptivity: the natural cycle as the optimal model, the\novarian stimulation cycle as suboptimal, and the refractory endometrium induced by the\ninsertion of an IUD as a negative control (for review see  Ruiz-Alonso  et al. , 2012 ). We performed a\n t -test and selected genes showing an absolute fold-change >3 and a\nfalse discovery rate <0.05. Three different statistical approaches were employed, the\nunion of the T-Rex gene list (GEPAS) ( http://gepas.bioinfo.cipf.es/ ) and the SAM gene list ( http://www.stat.stanford.edu/_tibs/SAM/ ), intersected with the multitest gene\nlist ( http://www.bioconductor.org/ ).\nMathematically, the approach can be written as: [T-Rex U SAM]Xmulttest.\nInitially, the ERA was created as a customised array containing 238 differentially\nexpressed genes that were coupled to a computational predictor able to identify the\ntranscriptomic profiles of proliferative (PRO), pre-receptive (PRE), receptive (R) or\npost-receptive (POST) endometrial samples, regardless of their histological appearance.\nThese 238 genes were presented to the scientific community in  Díaz-Gimeno  et al.  (2011) . But even more\nimportant than the genes implicated is the prediction algorithm, which enables combining the\nexpression of all 238 analysed genes to reach a consensus clinical diagnosis.\nTo test its accuracy and reproducibility, ERA was compared to standard histological methods\nin endometrial biopsies collected throughout the menstrual cycle (n = 128), and results were\nmeasured by the quadratic weighted Kappa index ( Diaz-Gimeno  et al. , 2013 ). For the accuracy study, biopsies were\ngrouped into two cohorts: the training set (n = 79) for ERA machine-learning training and\ndating, and a test set (n = 49) for comparison between histological and ERA dating. For the\nreproducibility study, seven women underwent one ERA test and a repeat test 29–40 months\nlater on the same day of their cycle. Concordance values following luteinising hormone (LH)\npeak were 0.618 (0.446–0.791) and 0.685 (0.545–0.824) for the two pathologists. Further, the\nKappa index for inter-observer variability (0.622; 0.435–0.839) was sub-optimal. ERA dating\nachieved a concordance of 0.922 (0.815–1.000) with LH peak. ERA test reproducibility in the\nindicated subgroup was consistent in all patients ( Diaz-Gimeno  et al. , 2013 ). These data provided robust indicators\nfor the utility of ERA.\n\nThe WOI lasts 30–36 hours and, depending on the patient, occurs between LH + 6 to LH + 9 in\nnatural cycles or from P + 4 to P + 7 in hormonal replacement therapy (HRT) cycles ( Rincon  et al. , 2018 ) ( Fig. 1 ).\nDiagram representing duration and timing of the window of implantation\n(WOI).  The WOI lasts approximately 30–36 hours and, depending on the patient,\noccurs between LH + 6 and LH + 9 in natural cycles or between P + 4 and P + 7 in\nhormonal replacement therapy (HRT) cycles.\nThe initial ERA proof of concept in Caucasian patients with recurrent implantation failure\n(RIF) was published in 2013 ( Ruiz-Alonso  et\nal. , 2013 ) in a prospective multicentre interventional clinical trial.\nOur hypothesis was that implantation failure of endometrial origin is not a pathology or an\nendometrial dysfunction (conditions that stigmatise a patient), but rather a failure to\nsynchronise the developing embryo with a patient’s individual WOI. The study group included\n85 patients with RIF (4.8 ± 2.0 previous failed cycles) and at least four total\nmorphologically high-grade embryos or blastocysts transferred and no other explanation for\nthe implantation failures. The control group was 25 patients. We detected that 25.9% of\npatients with RIF showed a displaced WOI (advanced or delayed), while only 12% of control\npatients had such displacement. Therefore, we concluded that one in four patients with RIF\nhave a displaced/asynchronous WOI. Our computational algorithm classified these patients as\nnon-receptive endometrium either pre- (84%) or post-receptive (16%), which was further\nverified by a second ERA test. We translated these genomic results to the clinic by\ntransferring embryo(s) according to the WOI of the individual patient, providing a\n‘personalised embryo transfer’ (pET) resulting in a 50.0% pregnancy rate (PR) and 38.5%\nimplantation rate (IR), similar to that of controls. These results suggested that normal\npregnancy and implantation rates may be achieved in patients with RIF of endometrial origin\nif synchrony between the embryo and receptive endometrium is accomplished ( Ruiz-Alonso  et al. , 2013 ).\nThis initial study was further validated by the report of a clinical case of successful pET\nafter seven previous failed IVF attempts (four with autologous oocytes and three with donor\noocytes) ( Ruiz-Alonso  et al. ,\n2014a ). The case report was soon complemented by a pilot study of 17 patients\nundergoing oocyte donation who experienced from 1 to 6 failed implantations (2.9 ± 2.1) with\nroutine embryo transfer (ET), but were subsequently treated with pET after diagnosis of\ntheir WOI. Results after pET showed that these patients (with up to six previous failures)\nreached a 60% clinical PR, while a 19% PR was achieved after routine ET in a non-receptive\nendometrium diagnosed by ERA ( Ruiz-Alonso  et\nal. , 2014a ).\nAfter these initial reports, independent groups started to publish their own data using ERA\nto guide pET in their clinical practice. In 2015, a retrospective study in an Indian\npopulation ( Mahajan, 2015 ) analysed data from\nthree different groups: patients with RIF, patients with one previous failed cycle, and\npatients with atrophic endometrium (<6 mm). Their results revealed that 27.5% of patients\nwith RIF had a displaced WOI, while only 15% of patients with one previous failure had a\ndisplacement (similar to our data published in 2013). After pET, the overall ongoing PR in\nthe RIF group was 42.4% and IR was 33%, which was similar to that in the group of patients\nwith one failure. This finding again suggested that results in patients with RIF can be\nnormalised after pET. Interestingly, the ERA test revealed displaced WOIs in 25% of those\nwith atrophic endometrium, but after pET their PR was 66.7% despite having an endometrial\nthickness <6 mm. Similar cases have been reported for unresponsive 4-mm endometrium\n( Cruz and Bellver, 2014 ). Intriguingly, in\npatients with congenital uterine abnormalities such as uterus didelphys and with previous\nfailed ETs, different endometrial receptivity status was found in each hemiuterus ( Carranza  et al. , 2018 ).\nIn 2017, a retrospective analysis of 50 patients with RIF assessed the impact of pET guided\nby ERA in a Japanese population ( Hashimoto\n et al. , 2017 ). Approximately 24% of patients in the RIF group\nhad a displaced WOI, but after pET they reached a 50% PR, similar to that reported in\nprevious studies. In 2019,  Hromadova  et\nal.  (2019)  reported similar findings in the Czech Republic.\nRetrospective data from 85 patients (74 RIF cases and 11 controls) revealed that 36.5% of\nRIF patients showed a displaced WOI and 69.2% became pregnant after performing pET guided by\nERA.  Ota  et al.  (2019) \npublished a case report of a Japanese patient who achieved pregnancy with pET guided by ERA\nafter 11 previous failed attempts.  Simrandeep and\nPadmaja (2019)  reported three severe cases of RIF in Indian patients; two of the\npatients had a previous ERA performed at a different centre, and the recommendation for pET\nfor a displaced WOI was not followed, resulting another failure. Once pET was implemented,\nsuccessful clinical pregnancies were achieved in both patients.\nWhile these studies indicate the outcomes for patients who received pET based on their WOI,\nwhat is the clinical outcome in patients in whom transfers occur outside of their WOI\naccording to ERA? Such data were collected in a study comparing the clinical outcome of pET\nin 205 receptive (R) patients versus embryo transfers performed in 52 non-receptive (NR)\npatients according to the ERA test. The clinical outcome was 23% PR and 13% IR after\ntransfer in the NR phase, with 0% ongoing pregnancy rate (OPR); in contrast, when pET was\nperformed based on the R phase, 60% PR, 45% IR and 74% OPR were achieved ( Ruiz-Alonso  et al. , 2014b ).\nHowever, other retrospective publications have not found statistical clinical differences\nin pET versus ET in patients with RIF ( Patel\n et al. , 2019 ). Tan  et al.  observed that when\nembryos were chromosomally analysed, a higher IR and OPR was observed in pET versus ET (66.7\nvs. 44.4% and 58.3 vs. 33.3%, respectively), but these differences were not statistically\nsignificant due to the small sample size ( Tan\n et al. , 2018 ). Some authors undertook a different approach to\nevaluate the clinical efficiency of ERA, using retrospective cohort studies comparing\npatients with an indication of ERA treated by pET to those without an ERA indication, and\nyielding similar clinical results between these groups ( Bassil  et al. , 2018 ;  Neves  et al. , 2019 ;  Cozzolino  et al. , 2020 ). We should bear in mind\nthat, until 2020, patients with indication for ERA were the most difficult cases with\nseveral previous failures, as no explanation was found for their RIF of endometrial origin\neven after a through infertility workup. Therefore, the fact that pET in this RIF population\nwas able to obtain similar clinical results to those in ‘control patients’ is confirmatory\nof previous results, due to improved outcomes for the most difficult patients.\nRecently, we explored the effectiveness of personalized embryo transfer guided by ERA\ncompared to frozen ET (FET) or fresh embryo transfer (ET) ( Simón  et al. , 2020 ). This prospective open label\nrandomised clinical trial (RCT) included 458 patients younger than 37 years undergoing IVF\nwith blastocyst transfer at their first appointment, across 16 reproductive centres from\nEurope, America and Asia, and involved 30 co-authors together with the support of the ERA\nRCT Consortium. Intention-to-treat analysis revealed comparable clinical outcomes across\ntransfer types; however, there was a significantly higher cumulative pregnancy rate (CPR) in\nthe pET group (93.6%) than in FET (79.7%) ( P  = 0.0005) and ET (80.7%)\ngroups ( P  = 0.0013). By per-protocol analysis, pET resulted in a 56.2%\nlive-birth (LB) rate after first embryo transfer compared to 42.4% for FET\n( P  = 0.09) and 45.7% for ET (45.7%,  P  = 0.17). After\n12 months, pET resulted in significantly higher cumulative LB rate (71.2%) compared to FET\n(55.4%,  P  = 0.04) and ET (48.9%,  P  = 0.003). pET also\nyielded significantly higher PR at the first embryo transfer (72.5%) compared to FET (54.3%,\n P  = 0.01) and ET (58.5%,  P  = 0.05). Similar outcomes\nwere observed for first-transfer IRs, which were 57.3% for pET versus 43.2%\n( P  = 0.03) and 38.6% ( P  = 0.004) for FET and ET,\nrespectively. All groups exhibited similar obstetrical, delivery type and neonatal outcomes.\nWhile the RCT experienced an unexpectedly high patient drop-out (observed, 50%; expected,\n30%), the per-protocol analysis comparing pET to FET and ET arms revealed significantly\nbetter cumulative LB rates, PR and IR. These findings support that using the ERA test at the\nfirst appointment to guide pET may have clinical benefit. Further, an independent RCT\ncomparing frozen blastocyst transfer using conventional timing versus timing guided by ERA\nis under way (ClinicalTrials.gov Identifier:  NCT03558399 ).\n\nDiscovery of the genes involved in endometrial receptivity has been challenging. The\nbackground presented above further encompasses that the sets of genes identified within\ndifferent transcriptomic studies differs due to differences in experimental designs, type of\narray initially used, sampling conditions, inclusion criteria, sample size, day of the cycle\nwhen biopsies were obtained and statistical analysis applied to the results, among other\nfactors. In sum, all the studies aiming to identify the physiological transcriptomic profile\nacross the menstrual cycle reached the same conclusion: it is possible to accurately\ncatalogue endometria at different stages based on their transcriptomic signatures,\nspecifically the identification of the WOI (see above: ‘A decade of basic research leading\nto the transcriptomic characterisation of the human endometrium’). Further, the\nmachine-learning predictors used to relate these gene signatures with clinical diagnosis\nhave differed. As an example, in our test, the core of the receptivity diagnosis is powered\nby 134 ERA genes, while the remaining genes target putative WOI displacements.\nSince the publication of our seminal paper identifying the transcriptomic signature of\nendometrial receptivity ( Díaz-Gimeno  et\nal. , 2011 ), six different companies have launched commercial endometrial\ntranscriptomic tests under different acronyms with different evidence. WinTest from INSERM\n( www.inserm.fr/en ) is based on 11 genes\ndetected using RT-qPCR, with four publications demonstrating transcriptomic and clinical\nconsistency ( Haouzi  et al. ,\n2009 ;  Haouzi, 2015 ;  Bissonnette  et al. , 2016 ;  Haouzi  et al. , 2021 ). ERPeak\nfrom Cooper Surgical (USA) ( https://fertility.coopersurgical.com/genomics/erpeak-endometrial-receptivity-test/ )\nand ERMap from IGLS (Spain) ( https://www.igls.net/es/services/mapa-de-receptividad-endometrial/ ) both use\n40 genes with RT-qPCR supported by the same paper ( Enciso  et al. , 2018 ). ERT based on 100 genes is commercially\navailable from Yikon (China) ( www.yikongenomics.com ) but has not been reported in a peer-reviewed\npublication. BeREADY from Competence Centre on Health Technologies Ltd (Estonia) ( https://beready.ccht.ee/ ) is based on 67\ngenes supported by one publication in collaboration with our group ( Altmäe  et al. , 2017 ). BioER from Bioarray (Spain)\n( https://bioarray.es/es/info/BioEr-TEST-DE-RECEPTIVIDAD-ENDOMETRIAL-60 ) is\nbased on 72 genes but has not been supported by a peer-reviewed report or proof-of-concept\nstudy.\nTranscriptomic signature differences have also been considered for endometrial pathologies.\nIn  Garcia-Velasco  et al. \n(2015) , we assessed the endometrial receptivity gene signature in patients with\ndifferent stages of endometriosis using the ERA test. We concluded that the WOI gene\nsignature does not vary significantly for patients with endometriosis, even considering\ndifferent stages, compared to controls. Our study also indicated that expression of the gene\nset was not modified by the presence or stage of endometriosis, but instead by the day of\nthe cycle when the biopsy was obtained. In contradiction to statements by the opponent, this\nis not a new finding since our group and others have consistently demonstrated that\nendometrial receptivity is not detrimental to embryo implantation in oocyte recipients with\nendometriosis, who have outcomes comparable to oocyte recipients without endometriosis\n( Diaz  et al. , 2000 ).\nDifferent candidate endometrial markers for endometriosis have been suggested, but whether\nthis is causal or merely consequent of endometriosis, or even whether this has any\nclinically relevant impact on human embryo implantation, has not been elucidated.\nFurthermore, oocytes from donors with endometriosis yield poorer PRs than those from donors\nwithout endometriosis when donated to otherwise healthy infertile women ( Simón  et al. , 1994 ), suggesting\nan embryonic factor is involved in poor prognosis of endometriosis patients.\n\nTechnology is rapidly evolving, and critics should update their knowledge at the same pace.\nMicroarray and PCR-based clinical tests are being replaced by NGS technology ( Lowe  et al. , 2017 ). In January\n2017, the ERA test was moved from microarray-based to NGS-based technology ( Clemente-Ciscar  et al. , 2018 ),\nas noted in subsequent diagnostic reports. Results of ERA in the RCT that began in October\n2013 and ended in November 2017 were reconfirmed by NGS technology ( Simón  et al. , 2020 ). Thus, transitioning to new\nplatforms as technology advances is a viable option.\n\nAn important point noted by the opponent is that bulk tissue analysis obtained from a\n“blind” endometrial biopsy may not be accurate enough to perform the ERA test. Instead, the\nauthor offers some guidance by quoting a computational deconvolution system that we\nco-developed ( Suhorutshenko  et al. ,\n2018 ) but is now outdated. The best possible technology currently available to\nchallenge the ERA test in bulk endometrial tissue in any part of the uterine cavity is\nsingle-cell RNA sequencing (scRNA-seq). scRNA-seq can promote understanding of how an organ\nor tissue is arranged at the single-cell level by blending biology and genetics with\nmathematics, new computational tools and pragmatism. Cells are isolated using microfluidic\ncircuits and nanodroplets, and the mRNA of every cell is sequenced separately. The spatial\ndistribution of RNA or translated proteins can then also be mapped within a tissue or organ\n( https://data.humancellatlas.org ).\nThis technology was chosen as the 2018 breakthrough of the year by  Science ,\nand its application in the human endometrium is no exception ( Wang  et al. , 2020 ).\nIn 2020, we reported the characterisation of the human endometrial transcriptome at a\nsingle-cell level, revealing cell-specific expression signatures across the menstrual cycle.\nFrom 29 healthy oocyte donors, we obtained and analysed 73 180 individual endometrial cells\nusing microfluidics (Fluidigm) or nanodroplets (10× Genomics) ( Wang  et al. , 2020 ). Employing canonical markers\nand highly differentially expressed genes, we identified six endometrial cell types:\nepithelial and endothelial cells, stromal fibroblasts, macrophages, lymphocytes and a novel\nciliated epithelial cell type. Further, the signatures revealed that the human WOI involves\ntranscriptomic activation in the epithelia that is both abrupt and discontinuous ( Figure 2 ) ( Wang  et al. , 2020 ). These cellular-resolution findings confirmed\nour previous identification from bulk tissue of a unique endometrial receptivity\ntranscriptomic signature ( Díaz-Gimeno  et\nal.  2011 ).\nTemporal transcriptome dynamics of endometrial transformation across the human\nmenstrual cycle by single-cell RNA sequencing (scRNA-seq).  The human WOI opens\nwith abrupt and discontinuous transcriptomic activation in the epithelia. Cells\n(columns) were ordered by pseudotime. Dashed lines: continuous transition. Solid lines:\nboundaries between four major phases. Reprinted from Wang et al., 2020 with permissions\nfrom Springer Nature. Copyright © 2020, The Author(s), under exclusive licence to\nSpringer Nature America, Inc. Please note that subsequent re-use of this figure is not\npermitted under this article's Open Access licence. Permission for re-use must be\nrequested from Springer Nature.\nThe timing of biopsy in relationship to the WOI is also questioned. First, in its\ndevelopment, ERA was compared to the previous gold standard histological methods (n = 128)\nand concordance against LH peak was superior to histology rating ( Diaz-Gimeno  et al. , 2013 ) (see above: ‘The\nendometrial receptivity analysis’). Second, we recommend that endometrial biopsies be\nobtained at LH + 7 or human chorionic gonadotropin (hCG)+7 in natural cycles or at P + 5\n(120 hours) in HRT cycles. This timing maximises the potential to find a receptive WOI, as\noccurs in 70% of patients analysed at this timing (see above: A decade of ERA clinical\napplication”). Notably, however, the prediction of receptive status within the range of\n4 days around the WOI is a major achievement of the ERA test, particularly identifying WOI\ndisplacements to guide pET. Some clinics and doctors have performed biopsies earlier or\nlater, and the percentage of receptive cases decreases but the prediction of the WOI is\nfeasible. Importantly, a confirmatory biopsy is not necessary because our algorithm can\npredict receptivity timing with high accuracy except in specific displacements (<10% of\ncases analysed). The consistency of the WOI prediction was challenged blindly in one patient\nthrough four different biopsies over four months ( Cho  et al. , 2018 ). After receiving the report for the first\nbiopsy with explicit instructions on how to proceed, the authors instead embarked on a\nseries of additional endometrial biopsies at various timings blinded to us, in opposition to\nthe original recommendation. Biopsies two, three and four all corroborated our initial\nfinding ( Stankewicz  et al. ,\n2018 ).\nMore important than the timing of biopsy is to ensure that endogenous P levels are <\n1 ng/mL within 24 hours before the administration of exogenous P in HRT cycles or at the day\nof hCG administration or LH peak in natural cycles. This step is done to avoid premature\nactivation of the P receptor, which will trigger the initiation of the endometrial\nreceptivity program. Our suggested standard endometrial preparation is HRT because this\napproach is consistent and reproducible. After menstruation, ovarian quiescence is confirmed\nby vaginal ultrasound evaluation and E 2  administration starting from the first or\nthe second day (in Europe, typically E 2  valerate at a dose of 6 mg/day or\nE 2  hemihydrate patches delivering 150 µg every 48 hours; in the United States,\noral estrace 200 mg three times daily; there are other possibilities depending on the\ngeographical availability of drugs). Sonographic evaluation and P assessment should be\nperformed 7–10 days after the initiation of endometrial E 2  preparation. When\na ≥ 6-mm trilaminar endometrium is observed with an endogenous P serum level < 1 ng/mL,\nexogenous P is administered at a dosage and route used by physician/clinic for a period of\n5 days (P + 5 or 120 hours). Then, the endometrial biopsy for the ERA test should be\nobtained. In Europe, typically we use vaginal micronised progesterone (or similar) at a dose\nof 400 mg/12 h; in the United States, 50 mg intramuscular progesterone daily (or similar) is\nused. The pET should always be performed using the same protocol as that used for the cycle\nin which the WOI was diagnosed by the ERA test.\n\nERA has never been presented independently of progesterone levels (see previous section and\n Simón  et al. , 2020 ).\nFurthermore, while the route of progesterone administration as well as the serum and tissue\nP levels are debatable, the activation of the progesterone receptor (PR) is not. PR (A and\nB) activation is the main driver of the molecular changes that determine the WOI and the\ninitiation of pregnancy. In a collaborative study ( von Grothusen  et al. , 2018 ), we challenged the ERA prediction\nability by blocking the action of P at the cellular level through the antiprogestogen\nmifepristone, which binds to PR. Mifepristone is approved in many countries for emergency\ncontraception and early first-trimester medical abortion. Indeed, a single dose of 200 mg\nmifepristone in the immediate postovulatory phase is sufficient to prevent pregnancy by\nrendering the endometrium refractory or non-receptive without interrupting the normal\nmenstrual cycle ( Gemzell-Danielsson  et\nal. , 1993 ,  1994 ). We\ndemonstrated that a single dose of mifepristone on Day 2 after the LH peak (LH + 2)\ncompletely ablates the receptive transcriptomic profile as assessed by the ERA test. Control\nsamples were all staged around receptive stage as would be clinically expected for LH + 7.\nTreatment samples were all categorised as non-receptive ( von Grothusen  et al. , 2018 ). Bioinformatic\npathway analysis yielded 60 differentially expressed genes within the ERA signature,\nresponsible for the inactivation of the PR and glucocorticoid receptor, consistent with\nmifepristone action. This finding further demonstrates the capacity of the ERA to identify\npharmacologically induced non-receptive endometrium through the blockade of PR ( von Grothusen  et al. , 2018 ).\n\nNotably, the opponent quotes an independent study ( Mahajan, 2015 ) as our own study to suggest that we contradict ourselves. He\nfurther supports his argument with complex statistical perspectives to make the point that\nRIF of endometrial origin is so rare that it should not even be treated; his recommendation\nis to keep trying all over again, pretending to obtain different results.\nRegardless of the opponent’s opinion, RIF of endometrial origin is recognised as a concern\nby all clinicians who transfer euploid embryos that ultimately fail to achieve pregnancy.\nThe ERA test was initially created to solve the problem of our most difficult patients,\nnamely RIF of endometrial origin (see above: ‘A decade of ERA clinical application’) that is\nestimated to be present in 10% of all IVF cycles ( Bellver and Simón, 2018 ). The RCT exploring, at the first appointment, the\ncost-effectiveness of this approach compared to FET or fresh ET has been published. Per\nprotocol analysis demonstrated that pET increases the IR at the first embryo transfer by\n14.1% (pp) versus FET ( P  = 0.03) and by 18.7% versus fresh ET\n( P  = 0.004). LB rates, while not statistically significant, were\nincreased by 13.8% versus FET and 10.5% versus fresh ET ( Simón  et al. , 2020 ). Thus, it is up to readers\nto consider if this approach is reasonable to use in all patients.\n\nCrucially, the opponent disproves of our recent RCT because the trial was planned for\npatients ≤ 37 years old at their first IVF cycle. He argues that such patients are not in\nneed of any additional diagnostic effort to improve clinical results, beyond iterative\ntreatments. We leave it to readers to decide whether there is any room for improvement that\nwill be welcome in this group of patients.\n\nOn the cryopreservation of embryos, we strongly disagree with the opponent. Embryo\ncryopreservation is a consolidated technology that was initially created to store\nsupernumerary embryos, but ultimately changed IVF clinical practice worldwide. Many clinics\nare now free of ovarian hyperstimulation syndrome thanks to oocyte/embryo cryopreservation\n( Devroey  et al. , 2011 ;\n Griesinger  et al. , 2011 );\nas well as fertility preservation is possible in young women ( Donnez and Dolmans, 2013 ); and donor oocytes after storage closed\nsystem appear to produced normal obstetric and neonatal outcomes ( De Munck  et al. , 2016 ). A large multicentre\nrandomised trial assessed obstetrical and perinatal complications, congenital anomaly and\nneonatal death outcomes following transfer of either fresh or cryopreserved embryos among\n2157 women undergoing their first IVF cycle. These outcomes did not differ significantly\nbetween groups ( Table I ) ( Shi  et al. , 2018 ).\nThe incidence of obstetrical and perinatal complications, congenital anomaly and\nneonatal death in fresh embryo transfer compared to frozen embryo transfer groups.\nReproduced with permission from Shi  et al. ,  NEJM ,\n2018.\nNeonatal death was defined as the death of a newborn within 28 days after\ndelivery.\nThe fact is that out of 306 197 ART cycles performed at 456 reporting clinics in the United\nStates in 2018, resulting in 81 478 live-born infants, 103 078 were oocyte- or\nembryo-cryopreservation cycles in which all resulting oocytes or embryos were frozen for\nfuture use (Center for Disease Control and Prevention 2018 Fertility Clinic Success Rates\nReport). The same trend is observed worldwide except in countries where legislation prevents\nit, such as UAE. Therefore, arguing a lack of safety of embryo cryopreservation or the use\nof HRT to justify not investigating the endometrial factor with the ERA test does not stand\nin 2021.\n\nAs physicians, we cannot sit back and ignore the consequences of accepting that failures\noccur more often than not. This attitude passes a message to our patients that the only way\nforward is to persevere with doing the same failed approach while expecting a different\nresult. The opponent lives in a unique country in which a patient can, without financial\nburden, try as many attempts as she (or her doctor) needs, but this is not common throughout\nthe rest of the world. In reality, after the first IVF failure, half of all patients will\nchange doctors. Additionally, the majority of patients in the United States whose health\ninsurance coverage would support a second IVF cycle do not seek further care after a failed\ntreatment ( Domar  et al. ,\n2018 ), and in countries where government sponsorship supports multiple IVF cycles,\none failed cycle leads a third of patients to discontinue treatment ( Brandes  et al. , 2009 ). Discontinuation is also\nthree times more likely among patients without IVF insurance coverage than those with IVF\ninsurance coverage ( Bedrick  et al. ,\n2019 ). In developing nations, a lack of access to financial support requires\npatients to self-pay for IVF treatment, which most often means investing their lifetime\nfinancial savings in a single treatment. These phenomena underscore the need to improve\noutcomes of the first IVF attempt.\nThe notion of ‘add-on’ was created to disprove any attempt to improve the status quo. This\nconcept pretends to ignore that our routine basal IVF results are poor and expensive. The\nnext step has been to group all of them in the same category regardless of their scientific\nevidence and/or clinical results. Every attempt to improve the status quo from unproven\nstrategies such as praying, scratching or immunological treatment, to others with supportive\nRCTs such as embryoscope, PGT-A or ERA are considered all the same. The ultimate concern is\nthe economic burden that imposes additional technological efforts to improve our results at\nthe first attempt, obviating the economic pitfall implied in repeating the same process all\nover again and expecting different results. Yet, add-on treatments should not be implemented\nwithout evidence for their benefit. Instead, it is crucial to consider and leverage all\nexisting evidence that may enable the first IVF treatment to be the best possible attempt:\nafter all, it may be their only chance. This approach also circumvents economic concerns, by\nproviding the best possible care from the start, rather than requiring a patient to undergo\nseveral costly failed cycles first. Any new evidence-based procedure that offers a ≥ 10%\nincrease in LBR with respect to routine IVF for ≤10% of the cost of a round of IVF should be\nseriously considered and/or discussed with the patient.\nAs with previous controversies in medical science, from heart transplants to test-tube\nbabies, attitudes have changed dramatically with time. Progress is historically achieved by\nthe eternal battle between ‘the guardians of faith’ who wish to maintain the status quo,\nremaining skeptical to any new medical advances even when there is ample room for\nimprovement, and the ‘visionaries’ who see new angles to address the lack of progress in a\ngiven field as an opportunity to improve the status quo. Progress is inevitable sooner\nrather than later.\n\nNo new data were generated or analysed in support of this research. The data collected for\nthis manuscript is available in the original papers referenced.\n\nM.R-A., D.V. and C.S. contributed to the conception and design of the study. M.R-A., D.V.,\nC.G., J.C. and C.S. contributed to the acquisition of data, drafting of the article and\ncritical review of the final draft. All the authors have approved the final version to be\npublished.\n\nThe authors declare no funding was given to this work.\n\nM.R., D.V., C.G. and J.C. are employees of Igenomix S.L. C.S. is co-inventor of the patent\nfor gene expression profile (ERA) issued to Igenomix and Head of the Scientific Advisory\nBoard of Igenomix.","source_license":"CC-BY-4.0","license_restricted":false}