Urinary gonadotropin assay on 24-h collections as a tool to detect early central puberty onset in girls: determination of predictive thresholds

Abstract

STUDY QUESTION Is the 24-h urinary gonadotropin assay an effective diagnostic tool in central precocious puberty (CPP) in girls? SUMMARY ANSWER This study is the first to provide 24-h urinary gonadotropin assay data, using an electrochemiluminescent immunoassay (CMIA), and to report its usefulness as a tool for the diagnosis of CPP. WHAT IS KNOWN ALREADY Data about the GnRH test in the diagnosis of CPP are variable and there is no consensus regarding its interpretation. The measurement of FSH and LH in urines was previously reported to be an alternative biological tool. STUDY DESIGN, SIZE, DURATION This is a retrospective two-cohort study, involving a setting and a validation cohort. A total of 516 girls, included between October 2012 and July 2015, and 632 urinary collections were analyzed in the setting cohort. In the validation cohort, 39 girls were included between January 2021 and May 2023, and 49 urinary collections were analyzed. PARTICIPANTS/MATERIALS, SETTING, METHODS This study included girls who consulted for an investigation of disturbed growth rate or a clinical suspicion of puberty onset in different medical centres across France (setting cohort). Girls with a suspicion of precocious puberty onset were addressed at the expert centre of paediatric endocrinology of the Groupement Hospitalier Lyon Est (validation cohort). Pelvic ultrasonography was performed and enabled their classification according to clinical and morphologic changes criteria (prepubertal or pubertal groups). The parents collected 24-h urine samples (u24) according to standardized instructions. FSH and LH (urinary or plasmatic) were measured using a current and automated CMIA. MAIN RESULTS AND THE ROLE OF CHANCE The area under the ROC curves for CPP prediction was 0.709 for u24FSH (P < 0.001), 0.767 for u24LH (P < 0.001), and 0.753 for the u24LH/u24FSH ratio (P < 0.001). We retained all possible combinations of the four thresholds in the validation cohort (u24FSH = 1.1 or 2.0 IU/24 h; u24LH = 0.035 or 0.08 IU/24 h). The combination of u24FSH > 1.1 IU/24 h and u24LH > 0.08 IU/24 h had a positive PV of 85.7% and a negative PV of 94.3%, a sensitivity of 85.7% and a specificity of 94.3%, for classifying prepubertal and pubertal girls in this cohort. LIMITATIONS, REASONS FOR CAUTION This is a retrospective study, in which a margin of error remains due to the inherent uncertainty regarding the clinical assessment of pubertal onset. It must be considered that the thresholds can only apply to the used reagents; measurements without extractions using other reagents are likely to show important heterogeneity. WIDER IMPLICATIONS OF THE FINDINGS The assay performed herein is a simple, non-invasive, and analytically robust technique meeting the criteria for an alternative to the GnRH test which could be used to supplement its lack of sensitivity. STUDY FUNDING/COMPETING INTEREST(S) No specific funding was used. All authors declared no conflict of interest. TRIAL REGISTRATION NUMBER In-house #23-5214 registered study.