Testosterone (TTT) affects bone development and resorption. Androgens mediate osteoblast proliferation and differentiation and increase bone matrix production. TTT also effects growth factors such as TGF-beta and IGF-1 that may be important in osteoblast proliferation. Estradiol is a potent product through peripheral aromatization. This enzyme is active in bone and thus bone effects may be mediated through estrogens as in postmenopausal osteoporosis. LHRH therapy can reduce estradiol levels by almost 50% (ref 11). TTT loss may also effect calcitonin leading to resorption by decreasing the effect of endogenous calcitonin. The mechanisms of TTT on bone are therefore not well understood.

Jackson (12) found that over 70% of men with hip fractures following falls had hypogonadism. The relationship of hypogonadism and male osteoporosis has been known for decades but only recently has research looked at osteoporosis induced by treatment of prostate cancer.

Melton LJ 3rd 2003 published a population study reviewing 429 men who had orchiectomies from 1956-2000 and compared the fracture risk to the community and found a 3.5-fold increase in risk.

Daniell's (14) 1997 review showed that orchidectomy increased the fracture risk from 1% (no hormonal therapy) to 28% at 7 years and 48% at 9 years (Figure 2). A later report (2000) focused on the rate of BMD loss following 26 men. Overall there was a 4% loss in BMD in years 1 and 2 and 2% pa after year 4. By 18 months it was statistically significant compared to baseline.

A much larger study by Oefelein (15) with 181 patients on androgen ablation showed an actuarial 4% fracture rate (median duration of androgen ablation was 44 months) at 5 years and 20% at 10 years. (Combining the 8 patients at risk at 10 years in the Daniell and Oefelein papers yields a 35% crude fracture rate.) In the Oefelein (15) paper there were 9 fractures (5 hip, 4 extremity, 1 spine (one patient had 2 fractures)). Only 5 of the 9 were confirmed to have osteoporosis on BMD scan, the rest were osteopenic on plain X-ray. 62% of these patients had metastatic disease but the authors clearly distinguished osteoporotic from pathologic fractures. Slender white patients are at the greatest risk for AA associated fractures. The fracture rate seen by Oefelein (15) should be compared with men over 65 years who have an incidence of 0.5% for hip fractures from osteoporosis. This gives an odds ratio of about 5 for hip fractures. The risk of any fracture over the age of 70 is 1% giving an odds ratio also of 5.

Townsend (16) in a survey of 224 patients on LHRH treatment for a mean of 22 months showed a 9% fracture rate. Wei (11) studied BMD in 32 men beginning AA. 63% (5/8) had osteopenia / porosis before treatment. A significantly higher percentage, 88% (21/24), met these criteria after more than a year of AA. On the basis of regression analysis, an estimated 48 months of AAA would be necessary to develop BMD criteria for osteopenia in the lumbar spine for a man with average BMD at the initiation of therapy. This has implications for followup practices for men with normal baseline BMDs.

Oefelein (17) has also reported from 195 men that the 24 cases (9 of these cases fractured before initiation of AA) with skeletal fractures had a significantly worse overall survival from prostate cancer. This factor remains on multiple regression analysis to correct for patients with metastatic disease and nadir PSA (RR 7.4). Baseline testosterone levels were not reported and this may affect both cancer response and osteoporosis risk.

Berruti (18) followed 35 men with AA and found the mean gm/cm2 decreases at both the hip and spine at 1 year. BMD decreased 2-5% in 8 men, 5-10% in 8 men and >10% in 3 men for the spine (and 6, 6, 3 respectively for the hip). It should be noted that a few men even had unexpected rises in BMD.

Overall the data shows a 4% annual loss of BMD in the femoral neck (CI 0-7%) in the first 2 years of AA (then 1% annually) and 3% in the hip (CI 2-4%). If you take a bone loss of 3% per year each 2-3 years of AA will double the fracture risk (1 unit SD of T-score is about 8-10%). Added to this are over 60% of cases that are already osteopenic and 5-20% osteoporotic at baseline (19, 20).