Repeated implantation failure versus repeated implantation success: discrimination at a metabolomic level.

STUDY QUESTION: Is there any difference at the serum metabolic level between women with recurrent implantation failure (RIF) and women with recurrent implantation success (RIS) when undergoing in vitro fertilization (IVF)?

SUMMARY ANSWER: Eight metabolites, including valine, adipic acid, l-lysine, creatine, ornithine, glycerol, d-glucose and urea, were found to be significantly up-regulated in women with RIF when compared with women with RIS.

WHAT IS KNOWN ALREADY: Despite transfer of three high-grade embryos per cycle, RIF following three or more consecutive IVF attempts occurs in a group of infertile women. Conversely, there is a group of women who undergo successful implantation each cycle, yet have a poor obstetric history.

STUDY DESIGN, SIZE, DURATION: This study was conducted over a period of 10 years (January 2004-October 2014). Groups of 28 women with RIF (age ≤40 years and BMI ≤28) and 24 women with RIS (age and BMI matched) were selected from couples with primary infertility reporting at the Institute of Reproductive Medicine, Kolkata, India. Women recruited in the RIF group had history of implantation failure in at least three consecutive IVF attempts, in which three embryos of high-grade quality were transferred in each cycle.

PARTICIPANTS/MATERIALS, SETTING, METHODS: Blood samples were collected from both the groups during the implantation window following overnight fasting for at least 10 h (7-10 days post ovulation). Samples were analyzed using a 700 MHz NMR spectrometer and acquired spectra were subjected to chemometric and statistical analysis. Serum levels of endothelial nitric oxide synthase (eNOS) were measured using an enzyme immunoassay technique.

MAIN RESULTS AND THE ROLE OF CHANCE: Valine, adipic acid, l-lysine, creatine, ornithine, glycerol, d-glucose and urea were found to be significantly down-regulated in women with RIS when compared with those with RIF, with fold change values of 0.81, 0.82, 0.79, 0.80, 0.78, 0.68, 0.76 and 0.74, respectively. Further, serum eNOS was found to be significantly lower in women with RIF when compared with RIS (P < 0.05), indicating possible impairment in nitric oxide production. Metabolites, mostly related to energy metabolism, lipid metabolism and the arginine metabolic pathway were found to be considerably altered and are likely to be associated with the RIF phenomenon. However, the interplay between these molecules in RIF is complex and holds merit for further exploration.

LIMITATIONS, REASONS FOR CAUTION: In-depth studies of the arginine metabolic pathway in endometrial tissues seem necessary to validate our findings. A limitation of the present study is that the metabolic level changes, eNOS and nitric oxide levels have not been investigated in the endometrial tissues of the two groups of women. It would be interesting to investigate whether there exists a direct link between metabolic dysregulation and genetic factors that affects implantation in RIF women.

WIDER IMPLICATIONS OF THE FINDINGS: We speculate that tissue metabolomics can provide an improved understanding of the metabolic dysfunction associated with RIF. The identification of serum metabolic marker(s) in women with RIS may help with strategies of early therapeutic intervention, which may improve the chances of implantation significantly in women otherwise susceptible to IVF failure.

STUDY FUNDING/COMPETING INTERESTS: One of the authors, S.R.C. acknowledges the Council of Scientific and Industrial Research (CSIR), Government of India [No: 9/81(1228)/14, EMR-I] for financial support.