A positive ReceptivaDx result for BCL6 does not correlate with abnormal ERA results or decreased expression of receptivity-associated markers: two sides of the endometrial receptivity coin in fertility evaluation and treatment.

A total of 173 patients had concurrent ERA and ReceptivaDx BCL6 evaluation using the same tissue. One patient (BCL6-negative) had an inconclusive ERA result and was excluded from analysis. Forty patients were BCL6-positive, and 132 patients were BCL6-negative. Demographic and treatment-related parameters are shown in Table 1 . Patients who were BCL6-positive had a slightly higher mean number of prior implantation failure attempts compared to those who were BCL6-negative (p=0.03), but the mean number of total euploid embryos transferred that resulted in implantation failure was similar. Two patients (1.5%) in the BCL6-negative group, and three patients (7.5%) in the BCL6-positive group, had three or more failed euploid transfers prior to the biopsy. Age, BMI, the proportion of nulliparous women, and number of prior miscarriages were similar. ERA results by ReceptivaDx BCL6 status from the same endometrial specimen are shown in Table 2 . The distribution of ERA results was significantly different between BCL6-positive and -negative patients (p=0.02). Contrary to our hypothesis, BCL6-negative patients had a higher proportion of pre-receptive result on ERA (22.0%) compared to BCL6-positive patients (2.5%) (p<0.01). Univariable logistic regression analyses were used to assess for potential association of various patient and treatment-related factors with a pre-receptive ERA result. A positive BCL6 result on ReceptivaDx was associated with reduced odds of a pre-receptive result on ERA (OR 0.09 [95%CI 0.01–0.69], p=0.02) ( Table 3 ). There were no statistically significant associations between age, BMI, nulliparity, endometrial preparation method, number of previous miscarriages, number of prior implantation failures, or number of prior unsuccessful embryos transferred and a pre-receptive ERA result. Our pilot single-nucleus sequencing study included four prospectively collected endometrial biopsy samples sent for BCL6 evaluation (ReceptivaDx ™ ): one natural cycle (NC) BCL6-positive endometrium; one NC BCL6-negative endometrium; one programmed cycle (PC) BCL6-positive endometrium; one PC BCL6-negative endometrium. Details of baseline patient characteristics, BCL6 HSCORE by ReceptivaDx, and endometrial preparation method for these four samples were described in Methods (and summarized in Supplemental Data Table S1 ). Both PC endometrium had concurrent ERA testing with a “receptive” result. Patients in the NC group did not undergo concurrent ERA evaluation. The four samples yielded 11,227 single nuclei for analysis after quality control filters. We identified 12 cell clusters over 5 major cell types ( Figure 1A ): epithelial cells (unciliated and ciliated), stromal fibroblasts, lymphoid cells, myeloid cells, and endothelial cells. Of these, the 5,718 single nuclei from epithelial cells were separated out for further analyses given the focus on epithelial receptivity-associated markers: 1,103 epithelial nuclei from NC BCL6-positive, 1,366 from NC BCL6-negative, 1,308 from PC BCL6-positive, and 1,941 from PC BCL6-negative endometrium. We compared epithelial cell transcriptomes from BCL6-positive to -negative endometrium within each endometrial preparation method group to limit potential differences resulting from exogenous hormones ( 21 , 22 ). Concordant with the epithelial BCL6 protein expression reported by ReceptivaDx, epithelial nuclei from a ReceptivaDx BCL6-positive endometrium had a significantly higher relative expression of BCL6 mRNA compared to those from a BCL6-negative endometrium in both NC and PC settings ( Supplemental Figure S1 ). UMAP presentation showed separation between BCL6-positive and BCL6-negative epithelial nuclei transcriptomes, particularly in NC endometrium ( Figure 1B ). Differential gene expression analysis of epithelial cell nuclei identified 2,801 significantly differentially expressed genes (DEGs; adjusted p-value <0.05) between BCL6-positive and -negative endometrium in the NC group (1,011 upregulated; 1,790 downregulated), and 1,062 DEGs in the PC group (703 upregulated; 359 downregulated). A total of 428 DEGs (230 upregulated; 198 downregulated) were shared across endometrial preparation methods ( Figure 1C ). Lastly, we examined relative expression levels of 34 receptivity-associated genes by BCL6 status in both epithelial cell nuclei from NC and PC endometrium ( Figure 1D ). These 34 genes were extracted from a published meta-analysis on the shared genomic signature associated with a receptive endometrium, and were reported to be upregulated specifically in epithelial cells ( 15 ). In NC endometrium, BCL6-positive epithelial nuclei transcriptomes showed significantly higher relative expression in 16 of the 34 receptivity-associated genes; the other 18 markers were not significantly differentially expressed. In PC endometrium, BCL6-positive epithelial nuclei transcriptomes showed significantly higher relative expression in 12 out of the 34 receptivity-associated genes; only one marker ( ANXA4 ) was significantly downregulated and 21 were not significantly differentially expressed.

A retrospective cohort study was performed on infertility patients at the University of California San Francisco. One hundred and seventy-three patients underwent endometrial biopsy using a Pipelle catheter (CooperSurgical, Trumbull, CT) for concurrent BCL6 (ReceptivaDx ™ ) and ERA ® (Igenomix) testing from 2019–2023. We only included patients with evaluation of both tests using the same endometrium to avoid individual cycle-to-cycle variability. Age, body mass index, prior pregnancy history, and number of prior implantation failures preceding endometrial evaluation were reviewed. Number of prior euploid embryo(s) transferred that resulted in failed implantation was also assessed. We defined implantation failure as having a negative serum hCG test after embryo transfer. Details regarding the endometrial preparation regimen, biopsy timing, and factors influencing the choice of endometrial preparation method were previously reported ( 13 ). None of the patients were on GnRH agonist or OCP treatment immediately preceding the endometrial biopsy cycle. Thirteen patients (7.5%) took letrozole for five days in the early proliferative phase prior to the biopsy. For this analysis, most patients underwent biopsy in a programmed cycle (PC) endometrium; all after 5 days of intramuscular progesterone-in-oil injections. A minority of patients underwent endometrial biopsy in a modified natural cycle endometrium, 7 days after choriogonadotropin alfa administration. Two patients were biopsied in a natural cycle (NC), both 7 days after endogenous LH surge. Neither ERA nor ReceptivaDx has published data on the utility/validation of their test stratified by endometrial preparation method (i.e., NC versus PC). Of note, BCL6 evaluation originated from studies in a natural cycle setting only. However, ReceptivaDx allows for (and does not differentiate) testing in either NC or PC settings. All samples were processed and sent to Igenomix and Cicero Diagnostics per instructions for ERA ® and ReceptivaDx ™ BCL6 testing, respectively. The primary outcome of the retrospective cohort study was the proportion of pre-receptive ERA result (to indicate a suboptimal profile of receptivity markers with potentially insufficient progesterone response) by ReceptivaDx BCL6 status. No prior studies compared these two tests directly for this primary outcome. Therefore, sample size calculation could not be accomplished a priori . An abnormal BCL6 result was defined as an HSCORE >1.4 as reported by ReceptivaDx ™ ; this threshold was used to dichotomize BCL6 status as “positive” or “negative”. ERA results were reported as “pre-receptive”, “receptive”, or “post-receptive”. We compared demographic and treatment-related parameters between BCL6-positive and -negative patients using t-test, chi-squared test, or Fisher’s exact test as appropriate. Differences in the proportion of pre-receptive ERA result by BCL6 positivity were assessed using the Fisher’s exact test. Logistic regression analyses were used to assess for an association between different patient/treatment-related factors and a pre-receptive ERA result. All analyses were performed at the 0.05 level of significance using Stata v17. A pilot study was conducted using single-nucleus transcriptomics to assess the relative transcriptomic levels of endometrial receptivity markers by ReceptivaDx BCL6 status. We prospectively collected endometrial biopsies using a Pipelle from four patients through our UCSF IRB-approved protocol (#10–02786). Biopsy samples were immediately flash-frozen in liquid nitrogen and stored at −80°C until single nuclei isolation. These four patients had unexplained infertility, a normal uterine cavity on saline sonogram, and no structural uterine pathologies (including polyps, fibroids, or adenomyosis) on imaging. Two patients were sampled in a natural cycle (NC), performed seven days after the LH surge. Two patients had a programmed cycle (PC) (estradiol patches 300mcg every 3 days, followed by intramuscular progesterone-in-oil injections 50mg daily) and endometrial sampling after five days of progesterone administration. In each group, one patient was positive, and the other negative, for BCL6 by ReceptivaDx. Additional information is summarized in Supplemental Data Table S1 . Single-nucleus sequencing was pursued as endometrial tissue was flash frozen while awaiting ERA and ReceptivaDx BCL6 determination; this approach allowed for efficient batch processing and sequencing of clinically relevant samples. Single nuclei were isolated from endometrial biopsy samples using the automated Singulator ™ 100 system (S2 Genomics, Inc.). Flash-frozen endometrial tissues were individually loaded for tissue dissociation and nuclei isolation. The cartridges were then placed into the Dissociation Chamber of the Singulator system. Nuclei isolation was performed using the Standard Nuclei Isolation V2 protocol, which follows the steps of tissue incubation, disruption, and washing with Nuclei Isolation Buffer and Nuclei Storage Buffers (Singulator ™ 100 system). Single nuclei suspensions were then washed with Nuclei Storage Buffer twice by centrifugation at 500g × 5min at 4°C followed by resuspension. The washed nuclei suspensions were passed through a 40uM Flowmi Cell Strainer ® to remove remaining cell debris and collected into a new tube. The nuclei concentration of each sample was determined using the Cellaca MX high-throughput automated cell counter (Nexcelom Bioscience, Inc.). An equal number of nuclei (n=1,000,000) from each sample was pooled together in preparation for single nuclei encapsulation. The pooled nuclei suspension was washed and resuspended in a 1x PBS solution containing 1% BSA and 0.2U/uL of RNase inhibitor. From the final pooled suspension, 40,000 nuclei were loaded onto the Chromium Controller for encapsulation and gel bead-in-emulsion (GEM) generation, followed by cDNA library construction at the UCSF Disease to Biology CoLab using the Chromium Single Cell 3’ Reagent v3.1 Kits (10x Genomics). The library was sequenced on the Illumina NovaSeq 6000 sequencer at a depth of 25,255 mean reads per nuclei. Cell Ranger (version 6.0.2) was used to align the raw sequencing data and generate a cell-by-gene matrix (default parameters with the addition of ‘--include-introns’). Using the filtered matrix from Cell Ranger, data were loaded into Seurat (v3) ( 16 ). Quality control was performed by filtering out nuclei with ≤300 gene counts (low-quality nuclei or empty droplets), and those with >5% mitochondrial RNA content. After removing unwanted nuclei, we performed log normalization and scaled the data, regressing out heterogeneity associated with cell cycle stage and percent mitochondrial contamination. Freemuxlet ( 17 ) was used to perform genetic demultiplexing of the pooled dataset. Genotyping was performed with the Infinium Global Screening Array (GSA 3.0) followed by imputation using the TOPMed Imputation Server ( 18 – 20 ). The resulting variants were filtered to Rsq >0.3 and MAF >0.05. Variable genes were identified using the Seurat FindVariableFeatures() function and used as input for principal component analysis (PCA). The top 10 principal components were first used as input to Seurat FindNeighbors(), and louvain clustering was performed with Seurat FindClusters() at resolution 0.5. The Uniform Manifold Approximation and Project (UMAP) non-linear dimensional reduction technique was used to visualize the cell clusters. Marker genes for the cell clusters were identified using the Seurat FindAllMarkers() function using the negative binomial test. The cell type of each cluster was annotated based on cell marker genes reported from published literature ( 10 , 11 ). Based on the top markers, clusters with high expression of mitochondrial genes or non-specific expression were manually removed. Following this removal, we repeated variable gene selection, PCA (using 13 principal components), clustering, and re-ran UMAP to obtain the dataset for analysis. In this pilot study, we focused primarily on the endometrial epithelial cells given their expression of many of the currently proposed receptivity markers. The epithelial cell clusters were portioned out from the initial dataset for further analyses and clustering using PCA and UMAP. Differential gene expression (DEG) analysis was performed between epithelial cell nuclei from ReceptivaDx BCL6-positive and - negative endometrium (compared within NC and PC endometrial preparation settings separately) using a Wilcoxon rank-sum test (implemented in Seurat FindMarkers()) with Bonferroni multiple testing adjustment. DEGs with absolute log-fold change >0.1 and adjusted p-value < 0.05 were retained. In a meta-analysis by Altmäe et al. , 57 genes were proposed as the putative receptivity markers to represent the meta-signature of human endometrial receptivity. Thirty-nine of the 57 markers were additionally confirmed to be upregulated in a “receptive” endometrium using RNA-sequencing analysis with FACS-sorted cells; 34 of which were specifically upregulated in epithelial cells ( 15 ). These 34 receptivity-associated genes were selected for analysis, and a dot plot was used to visualize their relative expression in epithelial cells by ReceptivaDx BCL6 status.

The human endometrium is a unique tissue compartment that undergoes complex molecular changes and structural transformation throughout a menstrual cycle ( 10 , 11 ). In reproductive medicine, the quest continues for clinical markers that correspond to a receptive (and also, importantly, non-receptive) endometrium for embryo implantation. Essential factors that orchestrate a maternal-embryo interface conducive to successful pregnancy development during the peri-implantation window also remain largely a black box. BCL6 positivity is proposed as a marker of progesterone resistance ( 23 ), and in fertility treatment, marketed to detect endometrial inflammatory conditions that can impact implantation ( 19 ). Our findings suggest that BCL6 positivity does not clearly correlate with a suboptimal profile of receptivity-associated genes, some of which are progesterone-regulated. We first used ERA microarray results as a surrogate marker for progesterone response on a genome-wide, bulk tissue level at the time of anticipated embryo transfer. Although whether ERA reflects endometrial receptivity and improves embryo transfer outcomes is challenged by existing data ( 25 , 26 ), ERA designation of a receptive endometrium reflects a transcriptomic signature with progesterone effects based on its developmental methodology in healthy ovum donors and patients ( 7 ). A pre-receptive ERA result, without evidence that it directly correlates with progesterone resistance, may indicate absent/insufficient progesterone effect given the selection methodology of the ERA panel. As shown in Table 2 , BCL6 positivity does not correlate with a pre-receptive ERA result. Only one patient (2.5%) who was BCL6-positive had a pre-receptive result. On the contrary, the proportion of pre-receptive results was higher in BCL6-negative endometrium (22.0%), and BCL6 positivity was associated with reduced odds of having a pre-receptive ERA result in the same biopsy specimen (OR 0.09, 95%CI [0.01–0.69], p=0.02). In fact, BCL6 is a part of the gene list used to construct the ERA, and was upregulated in what was proposed to be a receptive endometrium ( 7 ). Therefore, the reduced odds of a pre-receptive ERA result with BCL6 overexpression may not be an incidental finding, and highlights the discordance of these two tests’ interpretation of “endometrial receptivity”. Based on the study by Evans-Hoeker et al. , BCL6 transcript levels were also slightly (but significantly) increased in mid-secretory phase in normal controls. One interpretation to explain this conundrum would be that it is the magnitude, not the presence, of increased BCL6 expression that indicates possible pathology that would benefit from intervention in fertility treatment. The HSCORE threshold of 1.4 for BCL6 may strongly indicate the presence of endometriosis, but is this the applicable threshold to guide interventions in fertility treatment? Do all patients with endometriosis have clinically significant progesterone resistance that affects fertility (and benefit from medical/surgical treatment)? Non-randomized studies demonstrated encouraging results with improved pregnancy rates for BCL6-positive patients after medical/surgical treatment. A pilot randomized controlled trial in ten patients ( NCT04039204 ) comparing treatment to no-treatment for BCL6-positive patients with a history of prior failed euploid embryo transfer(s) also showed some preliminary positive results ( 4 , 5 ). To gain better resolution on the expression of receptivity markers, many of which are expressed by endometrial epithelial cells, we performed single-nucleus sequencing on four window of implantation endometrial biopsies. Of the 11,227 high-quality nuclei transcriptomes, 5,718 (50.9%) were from epithelial cells. BCL6 positivity correlated to higher transcript levels of BCL6 in epithelial cell nuclei within each endometrial preparation method group. Concordant with our prior study ( 13 ), when analyzing the four samples altogether, BCL6 transcript levels are higher in NC endometrium compared to PC endometrium, regardless of BCL6 overexpression status ( Supplemental Figure S1 ). This suggests endometrial preparation method alone has a critical impact on BCL6 regulation, likely in setting of exogenous progesterone administration. Differential gene expression analysis revealed widespread transcriptomic changes associated with BCL6 positivity ( Figure 1C ). Furthermore, 428 DEGs were shared in both NC and PC BCL6-positive endometrium, indicating possible common biological pathway(s) affected with BCL6 overexpression independent of endometrial preparation method. Gene ontology analysis using the shared DEGs (both upregulated and downregulated) revealed significant enrichment in genes involved in epithelial cell development, muscle contraction, cell communication by electrical coupling, and other pathways ( Supplemental Data Figure S2 ). We did not identify inflammatory-related pathways using the shared epithelial DEGs in this dataset. DEG and gene ontology results require confirmation with additional data from a larger sample size and analysis of additional cell types. Given the association between BCL6 positivity and endometriosis, additional investigation on the relationship between a positive ReceptivaDx result and endometrial inflammatory pathways will be valuable. In this pilot study, 34 receptivity-associated epithelial genes did not show a clear pattern of downregulation with BCL6 overexpression. Interestingly, most of these markers had higher relative expression in a BCL6-positive endometrium ( Figure 1D ). However, it is worth noting that the current canonical receptivity marker signatures also have questionable value in the clinical setting in prognosticating implantation and pregnancy outcomes. Together, our observations show that a positive ReceptivaDx BCL6 result does not correlate with a suboptimal profile of canonical receptivity markers by ERA and single-nucleus sequencing of epithelial nuclei. It is plausible that BCL6 positivity could still correlate to a deficient endometrial microenvironment (e.g., via endometriosis-associated inflammation), but its effect may not be reflected by evaluating receptivity markers, or fully manifest until after the window of implantation. Another possibility is the threshold used for BCL6 dichotomization (HSCORE >1.4) may not clinically differentiate progesterone resistance or suboptimal endometrial receptivity, although a higher BCL6 HSCORE (either as a continuous variable or categories of HSCORE >1.4) also was not statistically significantly associated with a history of failed implantation in this dataset ( Supplemental Table S2 ). There is likely a redundancy of signaling pathways in the development of a receptive endometrium, and the two commercial tests may address different portions of the multiplicity of pathways in the secretory phase. Strengths of our study include a large sample size of patients who had concurrent ReceptivaDx BCL6 evaluation and ERA microarray for correlation. This methodology avoids individual cycle-to-cycle variability that may affect the outcome of either result depending on endometrial preparation method for BCL6 evaluation ( 13 ), as well as, intra-patient variability that has been reported with ERA ( 27 ). Also, all patients were from a single center with overall homogeneous biopsy protocols. Both tests were performed at the respective third-party companies, reflecting what happens realistically in a clinical setting. The use of single-nucleus transcriptomics was novel, and all sequencing was performed in a single experimental run, thereby limiting batch effects. Our clinical findings are limited, however, by its retrospective nature, which we aimed to augment using prospectively collected samples to evaluate the transcriptomic changes specifically with a positive ReceptivaDx BCL6 result. Single-nucleus sequencing was conducted on a small number of patient samples, but yielded a high number of individual nuclei transcriptomes that showed general consistency with the commercial ERA test result. Clinical pregnancy outcomes for patients in this study are not yet available for analysis. Notably, our center currently recommends intervention if a test is performed and the result is abnormal (i.e., adjust progesterone timing if non-receptive ERA and primarily medical treatment following a positive BCL6 result). Therefore, a “no treatment” comparison group is unavailable to determine the outcome with no intervention. A recent publication by Lessey et al . reported higher pregnancy rates with GnRH agonist treatment compared to no treatment in sixty non-randomized patients; results of a pilot randomized controlled trial ( NCT04039204 ) with elagolix treatment was also reported by the authors ( 4 ). Notably, both ERA and single-nucleus RNA sequencing do not necessarily correlate perfectly with the proteome and functional state of the cells. Cell type-specific contributions and interactions in establishing endometrial receptivity are beyond the scope of this pilot study, and will be investigated with a larger sample size in the future. Lastly, the use of ERA in this study relies on the assumption that its gene panel signature reflects the presence of adequate progesterone response, which is logically deduced but not proven clinically. While an endometrial biopsy itself is relatively benign, current consequences of BCL6 testing in fertility treatment are significant. The currently proposed interventions with BCL6 positivity – hormonal suppression medically and/or laparoscopic surgery for endometriosis removal – carry significant burden, cost, and potential risks for patients. Therefore, additional studies are needed to confirm the utility and appropriate setting for BCL6 testing in fertility treatment. Data demonstrating the consistency of BCL6 as a prognostic marker of progesterone resistance from one menstrual cycle to another would be valuable. If BCL6 evaluation is pursued, the impact of different endometrial preparation methods, cycle-to-cycle variability, and medical/surgical treatment on BCL6 expression would provide insights to optimize treatment following an abnormal BCL6 result. Utilization of single-nucleus transcriptomics could aid our quest to understand the human endometrium – a unique, fascinating biological system that has proven itself difficult to comprehend.

BCL6 positivity, a proposed marker for progesterone resistance associated with impaired uterine receptivity, did not correlate with a pre-receptive ERA result. Receptivity-associated genes identified by a prior meta-analysis were not significantly downregulated in epithelial cells with single-nucleus sequencing of a BCL6-positive endometrium. These findings demonstrate discordance between the pragmatic interpretation of endometrial receptivity by ReceptivaDx and ERA, and highlight the need for further validation of endometrial evaluation methods in infertility treatment.

Failure of a good-quality embryo to implant in a normal-appearing endometrium remains an enigma in assisted reproductive technology (ART). Overall implantation rates peak at approximately 65%, even for euploid embryos, with no proven interventions that consistently improve outcomes for patients experiencing implantation failure ( 1 ). Numerous studies aimed to identify specific markers or genomic signatures to prognosticate an endometrium’s potential for embryo implantation, and the mechanisms underlying failed implantation are incompletely understood. Herein, we investigated whether two clinically utilized commercial tests in ART with distinct endpoints but potentially intersecting mechanisms for endometrial receptivity correlate with each other. Specifically, we focused on a proposed marker of endometrial inflammation and abnormal progesterone signaling (ReceptivaDx ™ , Cicero Dx, Inc) and a bulk transcriptome-based assay that aims to predict the window of implantation (Endometrial Receptivity Analysis (ERA ® ; Igenomix). ReceptivaDx ™ evaluates the expression of B-cell lymphoma 6 (BCL6) in the endometrium. Its output is a histologic score (HSCORE), which is a semiquantitative assessment of immunohistochemical staining of the BCL6 protein mainly in the endometrial epithelial cells ( 2 ). BCL6 overexpression in secretory-phase eutopic endometrium, proposed as a highly sensitive and specific marker for endometriosis, may also indicate the presence of other inflammatory conditions affecting the female reproductive tract ( 2 , 3 ). Therefore, BCL6 overexpression has been proposed as a marker of endometrial alterations associated with inflammation and progesterone resistance (as in endometriosis), and was shown in some studies to be associated with poorer IVF outcomes ( 4 – 6 ). ERA ® is a microarray tool that evaluates the profile of RNA transcripts extracted from bulk endometrial tissue. A total of 238 genes were used to develop the transcriptomic signature proposed to signify a receptive endometrium ( 7 ). This signature represents the differentially expressed genes when comparing bulk endometrial tissues from the time of luteinizing hormone surge + 7 days (LH+7) to endometrium from LH+1, LH+3, LH+5, and the proliferative phase. Whether ERA predicts implantation or improves embryo transfer success rates is contested and not discussed here. However, with its selection methodology, the differentially expressed genes (DEGs) in the ERA logically reflect the genome-wide downstream effects of progesterone response in healthy donors on a bulk-tissue level. A pre-receptive ERA result would indicate an endometrium in which its transcriptomic signature is more congruent with an early secretory endometrium that is outside the window of implantation and may not have fully initiated a response to progesterone. Furthermore, gene ontology analysis of the ERA genes also spans multiple biological processes across different cell types, including immune response, cell adhesion, and cell-cell signaling ( 7 ). Single-nucleus genomic approaches have accelerated rapidly in resolution over the past fifteen years. In bulk-tissue sequencing studies, gene expression changes by cell type and/or between conditions may be inadvertently missed due to background noise or dilutional effects from averaging gene expression values across whole tissue samples. In comparison, single-nucleus transcriptomics apply next-generation sequencing techniques to individual nuclei, thereby allowing cell type-specific resolution to deconvolute heterogeneous tissue structures and biological processes ( 8 ). Recent studies using this technology have provided new insights into the complex, dynamic nature of the human endometrium in both physiologic and pathologic states ( 9 – 11 ). We recently reported that endometrial preparation method correlates with BCL6 HSCORE (i.e., higher in natural cycle and lower in programmed cycle endometrium), despite no clear evidence showing one cycle preparation method is significantly superior over another clinically ( 12 , 13 ). This observation, along with another case-control study correlating livebirth outcomes with BCL6 expression, questions the clinical utility of BCL6 evaluation in the context of embryo transfer and fertility treatment ( 14 ). Endometrial receptivity in clinical practice is complex, poorly understood and prognosticated, and likely involves multiple facets and pathways. We aimed to evaluate the correlation between BCL6 positivity and one proposed aspect of endometrial receptivity: endometrial receptivity markers. We hypothesize that BCL6 positivity on ReceptivaDx, a proposed marker of clinically significant progesterone resistance associated with impaired uterine receptivity, correlates with a suboptimal profile of ERA and other receptivity-associated markers, many of which are progesterone-regulated. This may manifest as a pre-receptive transcriptomic signature on ERA, which was derived from early secretory endometrium outside the traditional window of implantation. Furthermore, we may also observe downregulation of receptivity-associated markers of epithelial cells, the first site of contact for implantation and a major cell type that expresses many currently proposed receptivity markers ( 15 ). In this study, we compared 1) the frequency of pre-receptive ERA result, and 2) relative gene expression levels of receptivity-associated markers in endometrial epithelial cells using single-nucleus RNA sequencing by ReceptivaDx BCL6 status.

Figure S1. BCL6 expression by endometrial preparation method and ReceptivaDx BCL6 status Figure S2. Gene ontology (GO) analysis on shared DEGs in epithelial nuclei. A) GO analysis using genes upregulated in BCL6-positive endometrium . B) GO analysis using genes downregulated in BCL6-positive endometrium. Figure S3: Venn diagram showing number of unique and shared significantly differentially expressed genes (adj. p<0.05) in stromal cell nuclei by ReceptivaDx BCL6 status in NC and PC endometrium. Figure S4: Venn diagram showing number of unique and shared significantly differentially expressed genes (adj. p<0.05) in immune cell nuclei by ReceptivaDx BCL6 status in NC and PC endometrium.