P–127 Does sperm quality really matter for conceiving through an intrauterine insemination program? A retrospective analysis of 5920 attempts taking into account repetition of cycles

Study question Are parameters of sperm quality part of the prognosis factors for an infertile couple to obtain a live birth when entering an intrauterine insemination program considering repetition of attempts? Summary answer Paradoxically, a lower sperm morphology independently predict quick live birth through IUI, as well as younger female age, lower D3 FSH, and higher triggering estradiol. What is known already Many studies have highlighted different prognosis factors for obtaining a live birth after IUI involving: male parameters (semen),female parameters (age, parity, ovulation, tubal, and endometriosis status, history of pelvic surgery),couple parameters (duration of infertility, number of previous attempt),IUI parameters (follicle number, endometrial thickness, estradiol at triggering, day of IUI, number of spermatozoa inseminated). However, most of these studies have included small number of attempts, semen parameters were either not collected or assessed with heterogeneity, and repetition of attempts (although iconic for IUI) was not considered, allocating inappropriate weight to cycles which failed in conceiving. Study design, size, duration We retrospectively studied the entire cohort of IUI attempts carried out with partner’s sperm at our center between 09/09/2003 and 01/17/2017. We included all male, female, couple and IUI parameters available. Each basic semen analyzes included have been carried out by a restricted number of skilled andrologists from our center. The closest semen assessment performed before IUI was considered as male parameter. IUI attempts were considered repeated unless a live birth or IUI abandonment occurred. Participants/materials, setting, methods Our primary outcome was live birth occurrence. We included 2228 couples having performed 5920 IUI attempts, with 636 live births obtained. A mixed logistic regression model was used to take into account IUI repetition before obtaining a live birth. A survival analysis using Cox model, with IUI rank as time variable, live birth as endpoint, and taking into account recurrences was carried out to determine which parameter best predict a quick live birth through IUI. Main results and the role of chance Included women were 33.7 ± 4.6 years old in mean. Baseline semen assessment was available for 64% of couples. Muti-variate analysis showed that live birth was more frequent when: Femal factors: age was young (33 to 38yo, OR 0.76 [0.60;0.96], >38 yo OR 0.49 [0.35;0.67], p = 0.0001), FSH ≤ 8.0 (OR 0.59 [0.45;0.79], p = 0.0002), AMH > 8.9ng/mL (OR 0.59 [0.45;0.79], p = 0.0001), endometriosis was absent (OR 0.56 [0.36;0.88], p = 0.0109), the patient already delivered (OR 1.36 [1.06;1.74], p = 0.0034) Male factors: sperm motility ≤26.0% (OR 0.71 [0.53;0.96], p = 0.0062), sperm vitality ≤72.0% (OR 0.65 [0.47;0.90], p = 0.0032), sperm typical form ≤ 25.0% (OR 0.51 [0.34 ; 0.78], p = 0.0016), IUI attempt factors: total dose of gonadotropin > 495.0 (OR 1.54 [1.26;1.88], p = 0.0001) follicle number > 2 (OR 1.23 [1.69;2.21], p = 0.0296), Estradiol at trigerring > 215.0pg/ml (OR 1.90 [1.53;2.36], p = 0.0001), Endometrial thickness >9.6mm (OR 1.43, p = 0.0024), day of IUI >13 (1.53 [1.24;1.89], p = 0.0001). Using Cox model, couples obtained quickly a livebirth if: woman age was below 33yo: For 33 to 38 yo, OR 0.37 [0.25;0.54], p = 0.0001, >38yo OR 0.19 [0.11;0.32], p = 0.0001 D3 FSH<8 (if above, OR 0.55 [0.34;0.90], p = 0.0160) - Sperm typical form ≤25.0 (if above, OR 0.34 [0.20;0.58], p = 0.0001) Triggering estradiol >215.0pg/ml, OR 1.99 [1.51;2.63], p = 0.0001) Limitations, reasons for caution Baseline semen assessment was missing a bit more when cycles were successful, so that a bias connot be excluded. This weird result concerning semen parameters which appear to lower the live birth rate when they are good should thus be considered with caution. Wider implications of the findings: Good baseline semen parameters do not appear as primordial for obtaining an IUI live birth, and were even found deleterious. However, thresholds highlighted in the present study were high, i.e. of limited clinical mean, and semen below them to remain normal. After confirming, explanations should be investigated: excessive acrosome reaction. Trial registration number Not applicable