TZD Use Is Linked to Increased Risk of Bone Fracture


NEW YORK — Treatment with a thiazolidinedione, either pioglitazone or rosiglitazone, was linked to an increased rate of bone fractures, particularly in women, in several recently published reports.

Although a definitive link between these drugs and an increased fracture risk has not yet been proved, the evidence amassed so far is suggestive enough to prompt caution in the treatment of patients with a thiazolidinedione (TZD), Dr. Robert G. Josse said at a meeting sponsored by the American Diabetes Association.

“In those with a higher fracture risk, consider other hypoglycemic therapy,” advised Dr. Josse, professor of medicine and nutritional sciences at the University of Toronto and medical director of the department of medicine at the osteoporosis center at St. Michael's Hospital in Toronto. In addition, “if using a TZD, consider therapy to prevent TZD-induced osteoporosis.” Standard therapies for osteoporosis are effective in patients with diabetes—including those with diabetes who develop steroid-induced osteoporosis—but no data now exist on the efficacy of antiosteoporosis treatments for countering the possible effects of TZDs, he noted.

Reasonable steps to reduce the fracture risk in patients who must take a TZD include optimizing calcium intake and the supply of vitamin D, encouraging adequate exercise, and taking precautions to prevent falls. Administration of antiresorptive drugs, such as raloxifene and the bisphosphonates, seems to be effective in patients with diabetes, but the effects of bone anabolic drugs such as teriparatide in these patients isn't known.

The idea that treatment with pioglitazone (Actos) or rosiglitazone (Avandia) may cause osteoporosis and produce an increased rate of bone fractures is biologically plausible, and has been suggested in the results from some studies.

Perhaps the most persuasive evidence so far is a meta-analysis published in January that compiled adverse-event data from 10 randomized, controlled studies with a total of more than 13,000 patients, and also reviewed two observational studies with a total of more than 31,000 patients (Can. Med. Ass. J. 2009;180:32–9). In the 10 randomized trials, patients treated with a TZD had a statistically significant 45% increased risk for bone fracture, compared with patients in the control groups. When the analysis broke the study population down by sex, a statistically significant 2.2-fold increased fracture risk was seen in women treated with a TZD, but absolutely no increased risk was seen in men. Additional analysis by sex showed that, in women, TZD treatment was linked with significant reductions of bone mineral density in the lumbar spine and hip. The two observational studies also showed a significant link between TZD use and fracture risk in women, but not in men.

The two short-term, randomized studies included a study with 50 healthy postmenopausal women without osteoporosis or diabetes who were randomized to treatment with 8 mg rosiglitazone daily or placebo for 14 weeks. Despite the brief period of treatment, the women in the rosiglitazone-treated group had a statistically significant reduction in their total hip bone mineral density, compared with the placebo group (J. Clin. Endocrinol. Metab. 2007;92:1305–10). A second study, published last May, randomized 30 postmenopausal women with polycystic ovary syndrome but without diabetes to treatment with either 30 mg pioglitazone daily or placebo. After 16 weeks, the women treated with pioglitazone had significantly lower lumbar spine and femoral neck density, compared with the controls (J. Clin. Endocrinol. Metab. 2008;93:1696–701). The TZD-treated women also showed significantly decreased blood levels of bone-turnover hormones and enzymes.

Dr. Josse reported receiving research support from, and serving on the speakers bureau and advisory panel for, several companies including Amgen Inc., Eli Lilly & Co., Procter & Gamble Co., and Sanofi-Aventis.

The evidence so far is suggestive enough to prompt caution in the treatment of patients with a thiazolidinedione. DR. JOSSE

Next Article: