Central Precocious Puberty in Girls: An Evidence-Based Diagnosis Tree to Predict Central Nervous System Abnormalities

Abstract

Objective. To identify predictors of central precocious puberty (CPP) that reveal central nervous system (CNS) abnormalities in girls with CPP.

Methods. A retrospective cohort study was conducted of all girls younger than 8 years with breast development related to CPP, seen between 1982 and 2000, in a university pediatric hospital in Paris, France. For a pilot population (186 idiopathic, 11 revealing CNS abnormalities), the accuracy of the Lawson Wilkins Pediatric Endocrine Society recommendations were evaluated. Potential clinical, radiological, and biological predictors of CNS abnormalities were assessed by univariate and multivariate analyses. A diagnosis tree aiming for 100% sensitivity for the detection of CNS abnormalities was constructed and was tested on a validation population (39 idiopathic, 3 revealing CNS abnormalities).

Results. Applying the Lawson Wilkins Pediatric Endocrine Society recommendations, 2 of 11 girls with CPP that revealed CNS abnormalities would not have been considered to require brain imaging. Independent predictors of CNS abnormalities were age at onset of puberty <6 years (adjusted odds ratio [AOR]: 6.7; 95% confidence interval [CI]: 1.5–29), lack of pubic hair at diagnosis (AOR: 7.7; 95% CI: 1.8–33), and estradiol >110 pmol/L (AOR: 4.1, 95% CI: 1.0–17). The diagnosis tree that was constructed on the basis of these predictors had 100% sensitivity and 56% specificity for the validation population.

Conclusion. The identification of girls who have CPP and require cerebral imaging seems possible on the basis of validated, simple, and reproducible predictors: age and estradiol. However, this process needs to be tested on other populations.