BACKGROUNDEmbryo transfer (ET) is a crucial step of in vitro fertilisation (IVF) treatment, and involves placing the embryo(s) in the woman's uterus. There is a negative association between endometrial wave-like activity (contractile activities) at the time of ET and clinical pregnancy, but no specific treatment is currently used in clinical practice to counteract their effects. Oxytocin is a hormone produced by the hypothalamus and released by the posterior pituitary. Its main role involves generating uterine contractions during and after childbirth. Atosiban is the best known oxytocin antagonist (and is also a vasopressin antagonist), and it is commonly used to delay premature labour by halting uterine contractions. Other oxytocin antagonists include barusiban, nolasiban, epelsiban, and retosiban. Administration of oxytocin antagonists around the time of ET has been proposed as a means to reduce uterine contractions that may interfere with embryo implantation. The intervention involves administering the medication before, during, or after the ET (or a combination).OBJECTIVESTo evaluate the effectiveness and safety of oxytocin antagonists around the time of ET in women undergoing assisted reproduction.SEARCH METHODSWe searched the Cochrane Gynaecology and Fertility (CGF) Group trials register, CENTRAL, MEDLINE, Embase, PsycINFO, CINAHL, and two trials registers in March 2021; and checked references and contacted study authors and experts in the field to identify additional studies.SELECTION CRITERIAWe included randomised controlled trials (RCTs) of the use of oxytocin antagonists for women undergoing ET, compared with the non-use of this intervention, the use of placebo, or the use of another similar drug.DATA COLLECTION AND ANALYSISWe used standard methodological procedures recommended by Cochrane. Primary review outcomes were live birth and miscarriage; secondary outcomes were clinical pregnancy and other adverse events.MAIN RESULTSWe included nine studies (including one comprising three separate trials, 3733 women analysed in total) investigating the role of three different oxytocin antagonists administered intravenously (atosiban), subcutaneously (barusiban), or orally (nolasiban). We found very low- to high-certainty evidence: the main limitations were serious risk of bias due to poor reporting of study methods, and serious or very serious imprecision. Intravenous atosiban versus normal saline or no intervention We are uncertain of the effect of intravenous atosiban on live birth rate (risk ratio (RR) 1.05, 95% confidence interval (CI) 0.88 to 1.24; 1 RCT, N = 800; low-certainty evidence). In a clinic with a live birth rate of 38% per cycle, the use of intravenous atosiban would be associated with a live birth rate ranging from 33.4% to 47.1%. We are uncertain whether intravenous atosiban influences miscarriage rate (RR 1.08, 95% CI 0.75 to 1.56; 5 RCTs, N = 1424; I² = 0%; very low-certainty evidence). In a clinic with a miscarriage rate of 7.2% per cycle, the use of intravenous atosiban would be associated with a miscarriage rate ranging from 5.4% to 11.2%. Intravenous atosiban may increase clinical pregnancy rate (RR 1.50, 95% CI 1.18 to 1.89; 7 RCTs, N = 1646; I² = 69%; low-certainty evidence), and we are uncertain whether multiple or ectopic pregnancy and other complication rates were influenced by the use of intravenous atosiban (very low-certainty evidence). Subcutaneous barusiban versus placebo One study investigated barusiban, but did not report on live birth or miscarriage. We are uncertain whether subcutaneous barusiban influences clinical pregnancy rate (RR 0.96, 95% CI 0.69 to 1.35; 1 RCT, N = 255; very low-certainty evidence). Trialists reported more mild to moderate injection site reactions with barusiban than with placebo, but there was no difference in severe reactions. They reported no serious drug reactions; and comparable neonatal outcome between groups. Oral nolasiban versus placebo Nolasiban does not increase live birth rate (RR 1.13, 95% CI 0.99 to 1.28; 3 RCTs, N = 1832; I² = 0%; high-certainty evidence). In a clinic with a live birth rate of 33% per cycle, the use of oral nolasiban would be associated with a live birth rate ranging from 32.7% to 42.2%. We are uncertain of the effect of oral nolasiban on miscarriage rate (RR 1.45, 95% CI 0.73 to 2.88; 3 RCTs, N = 1832; I² = 0%; low-certainty evidence). In a clinic with a miscarriage rate of 1.5% per cycle, the use of oral nolasiban would be associated with a miscarriage rate ranging from 1.1% to 4.3%. Oral nolasiban improves clinical pregnancy rate (RR 1.15, 95% CI 1.02 to 1.30; 3 RCTs, N = 1832; I² = 0%; high-certainty evidence), and probably does not increase multiple or ectopic pregnancy, or other complication rates (moderate-certainty evidence).AUTHORS' CONCLUSIONSWe are uncertain whether intravenous atosiban improves pregnancy outcomes for women undergoing assisted reproductive technology. This conclusion is based on currently available data from seven RCTs, which provided very low- to low-certainty evidence across studies. We could draw no clear conclusions about subcutaneous barusiban, based on limited data from one RCT. Further large well-designed RCTs reporting on live births and adverse clinical outcomes are still required to clarify the exact role of atosiban and barusiban before ET. Oral nolasiban appears to improve clinical pregnancy rate but not live birth rate, with an uncertain effect on miscarriage and adverse events. This conclusion is based on a phased study comprising three trials that provided low- to high-certainty evidence. Further large, well-designed RCTs, reporting on live births and adverse clinical outcomes, should focus on identifying the subgroups of women who are likely to benefit from this intervention.
