RESUMEN
OBJECTIVE: To study the impact of the quality of therapeutic control on fertility and on the prevalence of testicular adrenal rest tumours (TART) in young males with congenital adrenal hyperplasia (CAH). DESIGN: Combined cross-sectional and retrospective clinical study. METHODS: Twenty-nine patients and age-matched controls underwent clinical investigation, including semen analysis, testicular and adrenal ultrasound imaging, and serum and hair steroid analysis. The quality of therapeutic control was categorized as 'poor', 'moderate' or 'medium'. Evaluation of current control was based on concentrations of 17-hydroxy-progesterone and androstenedione in serum and 3 cm hair; previous control was categorized based on serum 17-hydroxy-progesterone concentrations during childhood and puberty, anthropometric and puberty data, bone age data and adrenal sizes. RESULTS: Semen quality was similar in males with CAH and controls (P = 0.066), however patients with 'poor' past control and large TART, or with 'poor' current CAH control had low sperm counts. Follicle-stimulating hormone was decreased, if current CAH control was 'poor' (1.8 ± 0.9 U/L; 'good': 3.9 ± 2.2 U/L); P = 0.015); luteinizing hormone was decreased if it was 'poor' (1.8 ± 0.9 U/L; P = 0.041) or 'moderate' (1.9 ± 0.6 U/L; 'good': 3.0 ± 1.3 U/L; P = 0.025). None of the males with 'good' past CAH control, 50% of those with 'moderate' past control and 80% with 'poor past control had bilateral TART. The prevalence of TART in males with severe (class null or A) CYP21A2 mutations was 53% and 25% and 0% in those with milder class B and C mutations, respectively. CONCLUSIONS: TART development is favoured by inadequate long-term hormonal control in CAH. Reduced semen quality may be associated with large TART. Gonadotropin suppression by adrenal androgen excess during the latest spermatogenic cycle may contribute to impairment of spermatogenesis.