IUI for unexplained infertility—a network meta-analysis

Abstract

ABSTRACT BACKGROUND IUI for unexplained infertility can be performed in a natural cycle or in combination with ovarian stimulation. A disadvantage of ovarian stimulation is an increased risk of multiple pregnancies with its inherent maternal and neonatal complication risks. Stimulation agents for ovarian stimulation are clomiphene citrate (CC), Letrozole or gonadotrophins. Although studies have compared two or three of these drugs to each other in IUI, they have never been compared to one another in one analysis. OBJECTIVE AND RATIONALE The objective of this network meta-analysis was to compare the effectiveness and safety of IUI with CC, Letrozole or gonadotrophins with each other and with natural cycle IUI. SEARCH METHODS We searched PubMed, MEDLINE, EMBASE, Cochrane Database of Systematic Reviews, CENTRAL and the Clinical Trial Registration Database indexed up to 16 August 2018. We included randomized controlled trials that compared a stimulation regimen with CC, Letrozole or gonadotrophins to each other or to natural cycle IUI among couples with unexplained infertility. We performed the network meta-analysis within a multivariate random effects model. OUTCOMES We identified 26 studies reporting on 5316 women. The relative risk (RR) for live birth/ongoing pregnancy rates comparing IUI with CC to natural cycle IUI was 1.05 (95% CI 0.63–1.77, low quality of evidence), while comparing IUI with Letrozole to natural cycle IUI was 1.15 (95% CI 0.63–2.08, low quality of evidence) and comparing IUI with gonadotrophins to natural cycle IUI was 1.46 (95% CI 0.92–2.30, low quality of evidence). The RR for live birth/ongoing pregnancy rates comparing gonadotrophins to CC was 1.39 (95% CI 1.09–1.76, moderate quality of evidence), comparing Letrozole to CC was 1.09 (95% CI 0.76–1.57, moderate quality of evidence) and comparing Letrozole to gonadotrophins was 0.79 (95% CI 0.54–1.15, moderate quality of evidence). We did not perform network meta-analysis on multiple pregnancy due to high inconsistency. Pairwise meta-analyses showed an RR for multiple pregnancy rates of 9.11(95% CI 1.18–70.32) comparing IUI with gonadotrophins to natural cycle IUI. There was no data available on multiple pregnancy rates following IUI with CC or Letrozole compared to natural cycle IUI. The RR for multiple pregnancy rates comparing gonadotrophins to CC was 1.42 (95% CI 0.68–2.97), comparing Letrozole to CC was 0.97 (95% CI 0.47–2.01) and comparing Letrozole to gonadotrophins was 0.29 (95% CI 0.14–0.58). In a meta-analysis among studies with adherence to strict cancellation criteria, the RR for live births/ongoing pregnancy rates comparing gonadotrophins to CC was 1.20 (95% CI 0.95–1.51) and the RR for multiple pregnancy rates comparing gonadotropins to CC was 0.80 (95% CI 0.38–1.68). WIDER IMPLICATIONS Based on low to moderate quality of evidence in this network meta-analysis, IUI with gonadotrophins ranked highest on live birth/ongoing pregnancy rates, but women undergoing this treatment protocol were also at risk for multiple pregnancies with high complication rates. IUI regimens with adherence to strict cancellation criteria led to an acceptable multiple pregnancy rate without compromising the effectiveness. Within a protocol with adherence to strict cancellation criteria, gonadotrophins seem to improve live birth/ongoing pregnancy rates compared to CC. We, therefore, suggest performing IUI with ovarian stimulation using gonadotrophins within a protocol that includes strict cancellation criteria. Obviously, this ignores the impact of costs and patients preference.