Prolactinomas in adolescents: persistent bone loss after 2 years of prolactin normalization

OBJECTIVE To evaluate the effect of hyperprolactinaemia and its treatment with dopamine-agonists on bone mass and turnover in adolescent patients compared to adults. PATIENTS Forty patients with hyperprolactinaemia (20 with disease onset during adolescence and 20 during adulthood) and 40 healthy control subjects. DESIGN Open transverse (in patients and controls) and open longitudinal (in the patients). MEASUREMENTS Bone mineral density (BMD) at lumbar spine and femoral neck, serum osteocalcin (OC) and urinary cross-linked N-telopeptides of type-1 collagen (Ntx) levels were evaluated in patients and controls. In the 40 patients, bone mass and turnover were re-evaluated after 12 and 24 months of treatment with bromocriptine (BRC, dose 2.5-10 mg daily), quinagolide (CV, dose 0.075-0.3 mg daily) or cabergoline (CAB, dose 0.5-1.5 mg weekly). RESULTS Transverse study: BMD values were significantly lower in hyperprolactinaemic patients than in controls, both at lumbar spine (0.81 +/- 0.01 vs. 1.010 +/- 0.01 g/cm(2); P < 0.001) and femoral neck (0.71 +/- 0.01 vs. 0.873 +/- 0.03 g/cm(2); P < 0.001). Thirty-two patients (80%) had osteoporosis and/or osteopenia at one or both skeletal sites. A significant inverse correlation was found between T score values measured at lumbar spine and femoral neck and the estimated disease duration. BMD was significantly lower in young than adult patients both at lumbar spine (T score, -2.4 +/- 0.1 vs. -1.4 +/- 0.3, P < 0.01) and at femoral neck (T score, -2.1 +/- 0.05 vs. -1.5 +/- 0.2, P < 0.05). Similarly, serum OC levels were significantly lower (2.0 +/- 0.11 vs. 9.1 +/- 2.4 mu g/l, P < 0.01) while Ntx levels were significantly higher in patients than in controls (129.2 +/- 1.7 vs. 80.7 +/- 2.9 nmol Bone collagen equivalent (BCE)/mmol creatinine; P < 0.001). A significant inverse correlation was found between prolactin (PRL) levels and OC levels, lumbar and femoral T score values, as well as between disease duration and OC levels, lumbar and femoral T score values. A significant direct correlation was also found between Ntx levels and PRL levels and disease duration. Longitudinal study: Normalization of serum PRL levels was obtained in all patients after 6-12 months of treatment. A significant increase of serum OC levels together with a significant decrease of Ntx levels was observed after 12 and 24 months of treatment (P < 0.01). Urinary and serum calcium, phosphorus, creatinine, and serum alkaline phosphatase and parathyroid hormone levels did not change during the study period in all patients. After 12 months of therapy OC and Ntx concentrations were restored to normal. A slight but not significant increase of BMD values was recorded after 12 and 24 months of treatment. After 12 months of treatment the percent increment of BMD values in the whole group of patients was 1.13 +/- 0.6% at lumbar spine and 1.2 +/- 0.4% at femoral neck level, whereas after 24 months, it was 2.8 +/- 0.7% at lumbar spine and 3.5 +/- 0.7% at femoral neck level. After 12 months of treatment, the percent increment of BMD values was 0.7 +/- 0.2% and 1.6 +/- 1.1% at lumbar spine and 0.9 +/- 0.5% and 1.6 +/- 0.5% at femoral neck level in the young and adult patients, respectively, whereas after 24 months, it was 2.1 +/- 0.8% and 3.4 +/- 1.3% at lumbar spine and 2.6 +/- 0.8% and 4.4 +/- 1.0% at femoral neck level in the young and adult patients, respectively. CONCLUSIONS Adolescents with prolactinoma have osteopenia or osteoporosis, a finding that strengthens the need for a prompt diagnosis. Since normalization of PRL concentrations by dopamine agonist therapy is unable to restore the bone mass, other therapeutic approaches should be considered in order to prevent further long-term problems.