SAN FRANCISCO – Evidence continues to mount that thiazolidinediones decrease bone density and increase fracture risk in patients with diabetes, but so far there are no randomized studies to show that adding an antiresorptive drug would protect bone health in this setting.
Bisphosphonates, denosumab, and parathyroid hormone have not been tested in human studies to see if they could reduce fracture risk in patients with diabetes who are taking thiazolidinediones, but randomized studies in mice suggest significant bone benefits from adding a bisphosphonate, Dr. Jonathan Graf said at a conference on osteoporosis sponsored by the University of California, San Francisco.
Dr. Jonathan Graf
For now, however, recommendations call for avoiding thiazolidinediones in patients with diabetes who have a high risk of fracture. A patient’s risk should be assessed by using the online FRAX tool and assessing the history for other risk factors for fracture that are not covered by FRAX, said Dr. Graf, a rheumatologist at San Francisco General Hospital.
If you do decide to use a thiazolidinedione to treat diabetes in someone at high risk for fracture, the detrimental effects on bone might be mitigated by using a lower dose. At least two thiazolidinediones – rosiglitazone and pioglitazone – are available in lower-dose formulations in combination with metformin, and the combinations provide the same level of diabetes control as higher-dose glitazone monotherapy, he said.
Preliminary data also suggest that combining lower doses of thiazolidinediones with an incretin drug may provide diabetes control while mitigating adverse effects on bone, he added. Incretins may even promote bone formation, some evidence suggests.
"There is not a lot of evidence out there to steer you one way or the other" in these management choices, but the evidence of damaging effects of thiazolidinediones on bone is becoming nearly overwhelming, he said.
In a prospective cohort study of 84,339 men and women with diabetes, the risk of fracture was 28% higher in patients on thiazolidinediones than in patients taking sulfonylureas, and the risk increased with cumulative exposure to thiazolidinediones. In men, pioglitazone increased fracture risk but rosiglitazone did not (Arch. Intern. Med. 2009;169:1395-1402).
A separate cohort study of 20,964 Medicare beneficiaries over age 65 years with diabetes found a statistical trend toward higher risk of peripheral fractures in both men and women on thiazolidinedione monotherapy compared with treatment with sulfonylureas or metformin (J. Clin. Endocrinol. Metab. 2009;94:2792-98).
Previous trials have suggested that the fracture risk is higher in women than in men. A meta-analysis of 10 trials and 2 observational studies containing a total of 45,484 patients with diabetes reported a doubling of fracture risk in women but not in men on thiazolidinediones, compared with controls (Can. Med. Assoc. J. 2009;180:32-9). In a separate analysis of a claims database, thiazolidinedione therapy doubled the risk of limb fracture in women (Am. J. Manag. Care 2009;15:491-6).
Thiazolidinedione therapy and the duration of therapy in older patients were associated with significant increases in nonvertebral fractures, including more than a fourfold increased risk of hip or femur fracture, a tripling in risk of forearm fracture, and a doubling in risk of humerus fracture, an observational study found (Arch. Intern. Med. 2008;168:820-5).
The 4-year fracture rate in patients on rosiglitazone doubled in both pre- and postmenopausal women but not in men compared with other diabetes treatments in A Diabetes Outcome Progression Trial (ADOPT), a randomized, double-blind study of about 3,600 patients. The fracture rate was 15% on rosiglitazone, 8% on glyburide, and 7% on metformin (Diab. Care 2008;31:845-51). At least three separate studies found significantly increased bone loss in women within weeks of starting thiazolidinediones, especially in postmenopausal women.
Thiazolidinediones are agonists to peroxisome proliferator-activated receptor gamma, which helps regulate bone formation. Thiazolidinediones decrease insulin-like growth factor 1 expression (which decreases bone formation), decrease osteoblastogenesis, and promote osteoblast differentiation.
"Just like glucocorticoids, they have several mechanisms that affect bone," he said.
Dr. Graf reported that he has no conflicts of interest.