Nervous system drugs are predicted as potential repurposing candidates for treatment in endometrial failure

Abstract Study question Is there a most represented group of approved drugs that can revert the impaired endometrial expression pattern associated with endometrial failure to improve reproductive outcomes? Summary answer Transcriptomics-based drug repurposing strategy identifies drugs currently used for nervous system diseases as repurposing candidates for endometrial therapies in infertility. What is known already Endometrium is a dynamic tissue with a key role in human reproduction whose alterations lead to infertility problems. Transcriptomic studies have been performed to understand molecular bases of endometrial failure, but effective treatments remain unknown. Transcriptomics-based drug repurposing strategy identifies approved drugs with a reversed disease gene expression profile, where genes up-regulated in a condition are down-regulated with the drug therapy and vice-versa, having a predicted therapeutic effect. Encouraging results for conditions such as preterm birth or cancer have already been shown using this methodology. In this study, we applied this approach to obtain treatment predictions for endometrial failure. Study design, size, duration A prospective multicenter study was performed between January 2019 and December 2021. A total of 192 women (18-50 years old) undergoing IVF with good-quality embryos were included and an endometrial sample in mid-secretory phase was collected to study their gene expression. Through artificial intelligence algorithms, we identified transcriptomic patterns with different endometrial prognoses. Expression changes among good or poor prognosis profiles were used to identify potential approved therapies to treat infertility leveraging drug expression databases. Participants/materials, setting, methods Gene signatures associated with poor prognosis were identified (FDR<0.05) and queried against the Connectivity Map database, containing gene expression profiles for 1,309 drugs in 5 different cell lines. Statistical analyses were performed to evaluate the drugs selected and obtain a score based on the drug expression reversal of the disease signatures. Significant drugs (adj. p-value) inversely associated with the gene signatures were highlighted. Finally, approved drugs were classified according to Anatomical Therapeutic Chemical codes. Main results and the role of chance We identified four transcriptomic profiles significantly (p-value<0.05) associated with different reproductive outcomes in the first embryo transfer after biopsy collection. Two poor prognosis profiles, one of them more related to clinical miscarriage, P1 (implantation rate = 30.4%, live birth rate = 42.9%, clinical miscarriage rate = 50%), and another to biochemical miscarriage, P2 (implantation rate = 20.0%, live birth rate = 33.3%, biochemical miscarriage rate = 66.7%) were compared with the best prognosis profile (implantation rate = 64.6%, live birth rate = 95.2%, biochemical miscarriage rate = 0.0%, clinical miscarriage rate = 4.8%), obtaining a total of 439 and 4,984 differential expressed genes, respectively for P1 and P2. After applying the drug repurposing methodology, we selected 50 and 32 significant approved drugs (adj. p-value<0.05) for P1 and P2. A total of 14 and 11 different drugs categories were obtained for each poor prognosis profile highlighting Nervous System (26%) and Alimentary tract and metabolism (12%) for P1, and Nervous System (23%), Antineoplastic and Immune modulating agents (15%) and Antiparasitic products (15%) for P2. Moreover, when both nervous system categories were compared, a total of 15 drugs were found in common, making possible the use of a unique drug for both profiles. Limitations, reasons for caution Since the Connectivity Map database does not include endometrial tissue, the action of these drugs should be validated experimentally in endometrial cell culture. Afterwards, to prioritize the best drug, side effects and approved doses reported in drug databases should be taken into consideration in further analyses. Wider implications of the findings The use of drug repurposing generates hypotheses for finding suitable drugs for endometrial failure, not requiring preclinical trials and going directly to Phase II clinical trials. These potential treatments will improve reproductive outcomes. Nervous system drugs seem to have a considerable effect on endometrium revealing possible causes of endometrial failure. Trial registration number Not Applicable