ESPE Abstracts

Background: Girls with Turner Syndrome (TS) are at an increased risk of primary ovarian failure. Oestrogen replacement is commenced at around the age of 12 years, in girls who do not enter puberty spontaneously, with incremental changes to the dose over the next 3 years until adult replacement doses are achieved. We have previously shown good correlation between serum and urinary LH (uLH) and FSH (uFSH) in children being assessed for disorders of puberty.

Objectives: 1. To determine whether there was a correlation between serum and urinary LH and FSH in hypergonadotrophic states;

2. To determine whether uFSH could similarly predict ovarian failure in TS as Anti-Müllerian Hormone (AMH).

Patients and Methods: A retrospective analysis of 37 TS girls attending the paediatric TS clinic in Glasgow between February 2015 and January 2019, in whom 96 non-timed spot urine samples were available with a median age at time of sample of 12.89 years (range 3.07-20.2 years). uLH and uFSH were measured by chemiluminescent microparticle immunoassay and corrected for urinary creatinine (uLHCr and uFSHCr). Simultaneous serum gonadotrophins and AMH were available in 30 and 26 girls, respectively. An AMH level<4 pmol/L was indicative of ovarian failure. Clinical information was collected from electronic case records and data were analysed using IBM SPSS (P<0.05).

Results: 48.6% had 45,X karyotype. 22/26 (84.6%) had started oestrogen replacement at a median age of 13.2 years (range 11.8-16.4 years). A strong correlation was found between serum LH and uLH (r 0.860, P<0.001) and serum FSH and uFSH (r 0.905, P<0.001). After correction for creatinine excretion the correlation remains significant, but the coefficient was smaller (uLHCr r 0.537, P=0.002 and uFSHCr r 0.381, P=0.038). Among patients ≥ 10 years not on oestrogen replacement, ROC curve identified uFSH as a reasonable predictor for AMH<4pmol/L: uFSH of >10.85 U/L predicts an AMH<4pmol/L with 75% sensitivity and 100 % specificity (AUC 0.875). uFSH had similar ability to predict ovarian failure in girls with TS (AMH <4pmol/L) as serum FSH (AUC 0.906).

Conclusion: uLH and uFSH are non-invasive, useful and reliable markers of ovarian insufficiency in hypergonadotropic states as TS. uFSH could provide an alternative to AMH (in centres which are limited by availability or cost) in predicting ovarian failure and the requirement for oestrogen replacement in pubertal induction.