Precocious puberty is defined as breast development before the age of 8 years of age.
• A problem is that an increasing percent of girls (especially in certain racial groups) are being classified as having precocious puberty using this criteria.
Central precocious puberty: due to activation of the hypothalamic-pituitary-gonadal axis (as opposed to ovarian tumor or other cause).
CNS abnormalities that might present with precocious puberty:
(1) astrocytoma or other glioma
(2) hamartoma
(3) subarachnoid cyst
Prevalence of CNS abnormality in study group: 6% (11 of 197 girls seen at a university pediatric hospital in Paris).
Algorithm for identifying a girl with central precocious puberty who should have CNS imaging studies:
(1) age < 6 years of age (19% CNS lesion)
(2) age >= 6 years of age, and serum estradiol (E2) > 45 pmol/L (4% chance CNS lesion)
A girl >= 6 years of age with a serum estradiol (E2) <= 45 pmol/L would not require imaging studies (0% risk of CNS lesion).
Additional notes:
(1) Chalumeau et al found that the lack of pubic hair at the time of diagnosis was a simple clinical predictor for CNS abnormality. Whether this finding will withstand further study is uncertain.
(2) The presence of headaches, seizures or other symptom related to the CNS, or rapid progression in pubertal development have been used by some to indicate girls who should have CNS imaging studies.
(3) The serum estradiol associated with CNS abnormality was > 110 pmol/L, but the cutoff point was reduced to be more sensitive.
Performance:
• 100% sensitive, 56% specific
Limitation:
• The normal reference range for serum estradiol was not listed. I could not find a Methods section to the paper. According to Tietz's Clinical Guide to Laboratory Tests, the reference range for a prepubertal girl should be < 55 pmol/L and for a girl in Tanner Stage 1 <= 37 pmol/L. It might be necessary to adjust the cutoff for local laboratory methodologies.
• A concern is that a rare girl with a detectable CNS lesion might be missed.
• As indicated by the authors, there is need for controlled studies to determine the most effective method of identifying those girls at significant risk for a CNS lesion.
• When tested in Brazil, the algorithm did not function as well, associated with a younger age of onset for puberty, more pubic hair and higher serum estradiol concentrations. Thus, the algorithm might need to be adjusted for different populations.
