L26/O-186 Clinical and ultrasound andrological parameters predictive for a pregnancy with a live birth

Abstract

Infertility affects 8%–12% of men. Clinical thresholds suggestive, but not predictive, of male infertility are commonly used, such as (i) seminal parameters < 5° centile of their distribution reported by the “WHO laboratory manual”, (ii) FSH > 8 U/L and low total or calculated free testosterone (with different cut-offs according to different scientific societies), (iii) mean testicular volume at Prader <15 mL and at ultrasound (US) < 12 mL. Conversely, only a few clinical cut-offs are available to predict male fertility, mainly related to semen parameters, usually defining “fertility” as the capacity to “induce a pregnancy” but not to “achieve a live birth”. This is a prospective, observational study with cross-sectional analysis of a first-time evaluation of clinical, seminal, hormonal, and male genital tract US characteristics of 604 males of infertile couples and 115 fertile men (derived from a spin-off of an US study on male fertility sponsored by the European Academy of Andrology) assessing differences between groups and investigating clinical and US predictors of a pregnancy with live birth. A total of 604 males (37.5 ± 7.0 years) of infertile couples and 115 males (36.6 ± 5.3 years) of fertile couples with live birth were evaluated in outpatient clinic and their parameters compared. For parameters with continuous distribution, iterative ROC curves identified thresholds predicting a pregnancy with live birth with the best sensitivity and specificity. These thresholds and other dichotomous parameters were used iteratively in binary logistic regression analyses adjusted for male and female partners age to evaluate the risk (odds ratio) to be men of a fertile couple. Comparing fertile and infertile couples, no difference in males or female partners age was observed. Evaluating clinical parameters, sperm concentration ≥ 40 × 106/mL, total count ≥ 128 × 106/ejaculate, progressive motility (PR) ≥ 52%, normal morphology ≥ 5%, FSH < 3.8 U/L, LH < 3.2 U/L, total and calculated free testosterone ≥17 and ≥0.335 nmol/L, respectively, mean testis volume (Prader) ≥ 20 mL, absence of epididymal dilation, and absence of cryptorchidism were predictive for a pregnancy with live birth. Evaluating US parameters, mean testis volume ≥ 16.5 mL, testis homogeneity, epididymis homogeneity, mean epididymal tail, and deferential ampulla <4.8 and 4.5 mm, respectively, prostate homogeneity and normal vascularization (defining absence of inflammation), absence of ejaculatory ducts dilation (defining absence of subobstruction), and seminal vesicles emptying > 2.8 mL (defining good seminal vesicles contraction) were predictive for a pregnancy with live birth. The absence of varicocele at US was not predictive of a pregnancy with live birth. Surprisingly, testis homogeneity at US showed the highest odds ratio for a pregnancy with live birth (OR = 33.7 [10.6–107.9]), followed by sperm concentration ≥ 40 × 106/mL and PR ≥ 52%. In a multivariate analysis adjusted for male and female partner age, including testis homogeneity, sperm concentration ≥ 40 × 106/mL, PR ≥ 52%, normal morphology > 5%, mean testis US volume > 16.5 mL, and FSH < 3.8 U/L, the best independent predictor of a pregnancy with live birth was testis US homogeneity (OR = 15.84 [4.58–54.8]). The thresholds found for clinical, seminal, hormonal, and US parameters can be used to predict, at first evaluation, if a male has characteristics predictive for a pregnancy with live birth. Scrotal US should be performed in outpatient clinics since testicular homogeneity represents the best predictor of pregnancy with live birth.