Abstract Study question What is the relative effectiveness and safety of final oocyte maturation trigger protocols in women undergoing in vitro fertilisation (IVF) treatment? Summary answer OHSS rate was lowest with GnRH agonist trigger and no difference in LBR was observed comparing various trigger protocols in normal, high and poor responders. What is known already Oocyte maturation trigger prior to oocyte retrieval is a crucial component of ovarian stimulation(OS). Pituitary suppression is an important component of OS in IVF, involving the use of GnRH analogues. Currently, the GnRH antagonist regimen is predominantly used given its comparable efficacy and lower risk of OHSS compared with GnRH agonist (GnRHa) regimens. Whilst only hCG trigger can be used with GnRH agonist regimens, GnRH antagonist regimen enables use of hCG trigger, GnRH agonist trigger or combinations of both as dual or double trigger. However, there is no consensus on how these trigger protocols compare in their effectiveness and safety. Study design, size, duration The following databases were searched until August 2022: MEDLINE, EMBASE, CINAHL, CENTRAL and ClinicalTrials.gov. Randomised controlled trials (RCTs) comparing at least two trigger protocols: hCG trigger, GnRH agonist trigger, dual trigger (hCG and GnRH agonist administered at the same time) and double trigger (GnRH agonist followed by hCG after a time interval) with the antagonist regimen were included. Primary outcome was live birth (LBR) per participant. Secondary outcomes included number of oocytes and OHSS rates. Participants/materials, setting, methods Two reviewers independently screened, selected studies and extracted data. Pairwise and network meta-analyses (NMA) were conducted according to ovarian response groups (normal, high and poor response). Effect estimates were presented as weighted means difference (WMD) and risk ratio (RR) with 95% confidence interval (CI) for continuous and dichotomous outcomes respectively. Quality assessment was performed using GRADE Main results and the role of chance Initial searches identified 4225 studies, of which 54 RCTs involving 5838 women met the inclusion criteria to be included in the analysis. In normal responders, there is no difference in LBR with the GnRHa vs hCG (RR:1.11,95%CI:0.83-1.49;3 studies,430 women,I2=33%,low-certainty evidence, direct comparison(DC)),dual trigger vs hCG(RR:1.14,95%CI:0.99-1.31;4 studies,1007 women,I2=33%,low-certainty evidence, DC),double trigger vs hCG(RR:0.53, 95%CI:0.27-1.06;indirect comparison(IC)),dual vs GnRHa (RR:0.97, 95%CI:0.70-1.34, IC),double vs GnRHa (RR:2.07, 95%CI:0.98-4.38, IC),dual vs double (RR:2.14, 95%CI:1.06-4.32;IC). In high responders, there is no difference in LBR with the GnRHa vs hCG (RR:1.04, 95%CI:0.84-1.29; 3 studies,178 women, I2 = 44%, low-certainty evidence, DC, dual trigger vs hCG (RR:1.82, 95%CI:1.25-2.67; 4 studies,117 women, I2 = 37%, low-certainty evidence, DC), double trigger vs hCG (RR:2.10, 95%CI:1.29-3.43, IC), dual vs GnRHa (RR:0.57, 95%CI:0.37-0.88, IC),double vs GnRHa (RR:0.49,95%CI:0.29-0.88, IC),dual vs double (RR:0.87, 95%CI:0.64-1.18; 1 study,57 women, low-certainty evidence,DC). In poor responders, there may be a difference in LBR when comparing dual trigger to hCG (RR:1.12,95%CI:1.12-2.89 1 study,112 women,low-certainty evidence,DC). OHSS rates were lowest with the use of GnRHa in high responders (RR:0.39,95%CI:0.04-3.78;9 studies;960 women;I2 =43%, low-certainty evidence,DC) and normal responders (RR:0.88, 95%CI:0.78 to 0.99;7 studies; 2246 women; I2=0%, low-certainty evidence,DC). There was no significant difference in number of oocytes or miscarriage risks with the use of any triggers. Limitations, reasons for caution Stratifying results by predicted ovarian response resulted in disconnected networks, limiting our ability to perform NMA for certain groups and outcomes. The certainty of the evidence was limited by high risk of bias. Wider implications of the findings Our results suggest that the use of short GnRH agonist trigger results in reduced OHSS rates in women with predicted normal or high ovarian response. There is no difference in LBR and number of oocytes comparing the different trigger protocols in all response groups (normal, high, poor responders) Trial registration number Not applicable
Read full abstract