Comparison of luteal support protocols in fresh IVF/ICSI cycles: a network meta-analysis

Despite the proven superiority of various luteal phase support protocols (LPS) over placebo in view of improved pregnancy rates in fresh cycles of IVF (in vitro fertilization) and ICSI (intracytoplasmic sperm injection) cycles, there is ongoing controversy over specific LPS protocol selection, dosage, and duration. The aim of the present study was to identify the optimal LPS under six core aspects of ART success, clinical pregnancy, live birth as primary outcomes and biochemical pregnancy, miscarriage, multiple pregnancy, ovarian hyperstimulation syndrome (OHSS) events as secondary outcomes. Twelve databases, namely Embase (OVID), MEDLINE (R) (OVID), GlobalHealth (Archive), GlobalHealth, Health and Psychosocial Instruments, Maternity & Infant Care Database (MIDIRS), APA PsycTests, ClinicalTrials.gov, HMIC Health Management Information Consortium, CENTRAL, Web of Science, Scopus and two prospective registers, MedRxiv, Research Square were searched from inception to Aug.1st, 2023, (PROSPERO Registration: CRD42022358986). Only Randomised Controlled Trials (RCTs) were included. Bayesian network meta-analysis (NMA) model was employed for outcome analysis, presenting fixed effects, odds ratios (ORs) with 95% credibility intervals (CrIs). Vaginal Progesterone (VP) was considered the reference LPS given its' clinical relevance. Seventy-six RCTs, comparing 22 interventions, and including 26,536 participants were included in the present NMA. Overall CiNeMa risk of bias was deemed moderate, and network inconsistency per outcome was deemed low (Multiple pregnancy χ²: 0.11, OHSS χ²: 0.26), moderate (Clinical Pregnancy: χ²: 7.02, Live birth χ²: 10.95, Biochemical pregnancy: χ²: 6.60, Miscarriage: χ²: 11.305). Combinatorial regimens, with subcutaneous GnRH-a (SCGnRH-a) on a vaginal progesterone base and oral oestrogen (OE) appeared to overall improve clinical pregnancy events; VP + OE + SCGnRH-a [OR 1.57 (95% CrI 1.11 to 2.22)], VP + SCGnRH-a [OR 1.28 (95% CrI 1.05 to 1.55)] as well as live pregnancy events, VP + OE + SCGnRH-a [OR 8.81 (95% CrI 2.35 to 39.1)], VP + SCGnRH-a [OR 1.76 (95% CrI 1.45 to 2.15)]. Equally, the progesterone free LPS, intramuscular human chorionic gonadotrophin, [OR 9.67 (95% CrI 2.34, 73.2)] was also found to increase live birth events, however was also associated with an increased probability of ovarian hyperstimulation, [OR 1.64 (95% CrI 0.75, 3.71)]. The combination of intramuscular and vaginal progesterone was associated with higher multiple pregnancy events, [OR 7.09 (95% CrI 2.49, 31.)]. Of all LPS protocols, VP + SC GnRH-a was found to significantly reduce miscarriage events, OR 0.54 (95% CrI 0.37 to 0.80). Subgroup analysis according to ovarian stimulation (OS) protocol revealed that the optimal LPS across both long and short OS, taking into account increase in live birth and reduction in miscarriage as well as OHSS events, was VP + SCGnRH-a, with an OR 2.89 [95% CrI 1.08, 2.96] and OR 2.84 [95% CrI 1.35, 6.26] respectively. Overall, NMA data suggest that combinatorial treatments, with the addition of SCGnRH-a on a VP base result in improved clinical pregnancy and live birth events in both GnRH-agonist and antagonist ovarian stimulation protocols.