Background and objective: Both adult- and childhood-onset GH-deficient adults are prone to osteopenia. Studies of GH replacement have, for the most part, demonstrated increases in bone mineral density (BMD). Previous studies have, however, used GH doses in excess of those currently used in low-dose titration regimens aimed at normalizing the serum IGF-I level. Furthermore, the effect of GH on the lumbar spine that has been irradiated during treatment of childhood cancer is unknown.
Patients: Thirty-two adult patients with childhood-onset GH deficiency were subdivided according to whether or not they had received spinal irradiation in childhood. The cohort in whom the spine had not been irradiated was comprised of 17 patients (seven male, 10 female), median age 29.8 years (range 20.6--40.8), the median age at primary pathological diagnosis being 9 years (range 4--16). The cohort who received spinal irradiation was composed of 15 patients (seven male, eight female), median age 22.9 years (range 16.5--40.3), with a median age at craniospinal irradiation of 9 years (range 2--16).
Measurements: At baseline, BMD was assessed at the lumbar spine and femoral neck by DXA, and at the ultradistal and distal radius by SXA. The patients were then commenced on GH replacement, titrating the dose at 4--6-weekly intervals to normalize the serum IGF-I level. BMD scans were reassessed following at least 1 year of GH replacement therapy. The mean duration of GH therapy was 1.68 plus minus 0.52 years.
Results: BMD was significantly reduced, compared to the reference data, at all four sites measured in both the spinally irradiated and unirradiated groups. No significant difference was observed between the subgroups with respect to lumbar spine BMD (P = 0.64). In the cohort who did not receive spinal irradiation, an increase in BMD of 3.5% above baseline was observed at the lumbar spine (P = 0.018), 13/17 patients showing a positive increment in lumbar spine BMD. No significant changes in BMD were observed at any other site within this cohort, or at any site in the patients who received spinal irradiation. A significantly greater change in lumbar spine BMD was observed in the unirradiated spine cohort compared with the spinally irradiated cohort (P = 0.018). No differences in response were demonstrated at the other sites studied between the two subgroups.
Conclusions: We have demonstrated an increase in lumbar spine BMD with long-term (mean 1.78 +/- 0.55 years) GH therapy in adults with childhood-onset GH deficiency, who have not received spinal irradiation, when GH was administered by a titration regimen aimed at normalizing the serum IGF-I level. No improvements were observed at the femoral neck, ultradistal or distal radius. Patients who had received spinal irradiation during childhood did not present with a reduction in spinal BMD in excess of that observed in the nonirradiated cohort. The spinally irradiated group were, however, resistant to the osteo-anabolic effects of GH, which we propose reflects the capacity of radiation-induced damage to suppress the skeletal response.