ESR1, ESR2 and FSH receptor gene polymorphisms in combination: a useful genetic tool for the prediction of poor responders.

PURPOSE: Previous studies in humans concluded that a multigenic model including specific FSHR, ESR1 and ESR2 genotype patterns may partially explain the poor response to FSH. The aim of our study is to analyse three different loci -polymorphisms in ESR1 Pvu II, ESR2 Rsa I and Ser680Asn FSH receptor gene- in a Greek population and their involvement in stimulation outcome and pregnancy rates.
METHODS: Each locus was studied alone, and in combination with the others. We performed both restriction fragment length polymorphism analysis and real-time polymerase chain reaction. A total of 109 normally ovulating female patients underwent IVF or ICSI.
RESULTS: Studying each locus alone, no significant results were drawn for ESR1 and ESR2 genes. Concerning the FSHR polymorphism, the women carrying the AA variant presented higher total amount of gonadotrophins used (P=0,048) and tended to have higher number of stimulation days (P=0,057). Considering the ESR1 and FSHR gene polymorphisms in combination, the TC/SA combination presents the highest number of pregnancies in poor responders group (3/4 pregnancies carried this genotype), in good responders group (4/12 pregnancies carried this genotype) and in the total population (10/26 pregnancies carried this genotype). Except the CC/AA combination, all other genotype combinations presented incidence of pregnancy, with TC/SA having the highest incidence. The CC/AA genotype presents the worst profile of ovulation induction, confirming a poor responder profile: the total amount of gonadotrophins used was highest in CC/AA group (P < 0,05). The peak E2, the number of follicles and of retrieved oocytes and the pregnancy rate were significantly lower (P < 0,05). This genotype combination seems to be over-presented in the poor responders group in a statistically significant way (P=0,038). Women with CC/AA combination have 1,5-2,4 times more risk to be poor responders in comparison with women that do not carry that combination.
CONCLUSION: This study supports the hypothesis that a multigenic model, including the well studied ESR1 and FSHR genes is involved in the controlled ovarian stimulation outcome indicating that the CC/AA genotype presents the worst ovulation induction profile, while the TC/SA genotype presents the higher number of pregnancies in our population.