Serum concentrations of type I and III procollagen propeptides in healthy children and girls with central precocious puberty during treatment with gonadotropin-releasing hormone analog and cyproterone acetate

Serum levels of type I and III procollagen propeptides (s-PICP and s-PIIINP) were measured in 466 healthy school children and in 23 girls with central precocious puberty (CPP) during GnRH analog and cyproterone acetate therapy, using two commercially available RIAs. In normal children, s-PICP and s-PIIINP changed significantly with age and pubertal development stages. For s-PIIINP, a peak was seen at 12 yr for girls and 13 yr for boys; no peak could be discerned for s-PICP. The prepubertal (Tanner stage 1) s-PICP value (mean +/- SD) for girls was 374 +/- 132 micrograms/L, the midpubertal value (stage 3) was 442 +/- 135 micrograms/L, and the postpubertal value (stage 5) was 203 +/- 103 micrograms/L. The mean s-PIIINP levels for girls were 9.1 +/- 2.4, 15.0 +/- 4.3, and 6.8 +/- 3.1 micrograms/L, respectively. For boys, levels were 362 +/- 119, 544 +/- 138, and 359 +/- 256 micrograms/L for s-PICP and 8.5 +/- 2.2, 14.5 +/- 5.0, and 8.6 +/- 3.8 micrograms/L for s-PIIINP (P < 0.001 for both propeptides in both boys and girls). There was, however, a large variation in normal values for both propeptides within the age groups and pubertal stages. There was a significant correlation of s-PICP and s-PIIINP levels to height velocity in girls (r = 0.35; P < 0.001 and r = 0.33; P < 0.001, respectively), while in boys, only s-PIIINP showed significant correlation to height velocity (r = 0.40; P < 0.001). In untreated girls with CPP, serum levels of s-PIIINP were elevated [PIIINP SD score (SDS), 2.13]. Levels of s-PICP were normal (PICP SDS, 0.39). Levels of both propeptides decreased within 2 months after initiation of therapy and remained below initial values (P < 0.01). The decrease in s-PIIINP after 2 months of therapy showed a significant correlation with the fall in height velocity SDS for chronological age after 6 months of therapy (r = 0.64; P < 0.01). We conclude that s-PIIINP and, to a lesser degree, s-PICP reflect growth in normal children, but due to the large variation, both propeptides seem unsuitable as markers for screening of growth disorders in children.