Conference Coverage

Bisphosphonate holiday may help to reduce atypical femur fracture risk



Women who take a 3- to 5-year break from bisphosphonates after 3-5 years on treatment could significantly reduce their risk of atypical femur fractures, according to a study of more than 150,000 Southern California health plan members presented at the annual meeting of the American Society for Bone and Mineral Research in Montreal.

Dr. Annette L. Adams of Kaiser Permanente Southern California

Dr. Annette L. Adams

A second study in the same group of women found that higher bone mineral density levels before bisphosphonate treatment may put women at greater risk for these fractures.

“Our findings suggest really strongly that among women who have achieved some level of bisphosphonate exposure a drug holiday is certainly a reasonable approach to trying to balance the risk of these atypical fractures versus the benefit of preventing more typical, or classic, hip fractures,” said presenter Annette L. Adams, PhD, MPH, a research scientist with Kaiser Permanente Southern California, who was an author on both studies.

Atypical femur fractures (AFFs) occur below the trochanters in the femoral shaft area, around mid-thigh, Dr. Adams said. They appear different from other fractures, usually breaking in just two pieces transversely through the center of the bone, and occur with minimal to no trauma, she said.

“We’ve heard stories of women that are just sitting in a chair, and they stand up and their femur fractures,” Dr. Adams said. “Or they’re out in the garden on their knees, and they try to stand up and it fractures. These are not necessarily fall-related like many traditional hip fractures.”

The researchers observed a 44% reduction in women’s risk of AFFs in the first year of a so-called drug holiday, or discontinuation of bisphosphonates, after 3-5 years on treatment when compared with those who stayed on the medications (hazard ratio = 0.56; 95% confidence interval, 0.38-0.82). During the first to fourth years after discontinuation, AFF risk was decreased by 80% (HR = 0.20; 95% CI, 0.10-0.37), and after 4 years, AFF risk was reduced by 78% (HR = 0.22; 95% CI, 0.08-0.59) in comparison to bisphosphonate users.

Dr. Adams and her colleagues reviewed records from 152,934 women aged 50 or older who were members of Kaiser Permanente Southern California between Jan. 1, 2007, and Sept. 30, 2015. There were 185 AFFs overall (incidence rate 1.70 per 10,000 person-years).

The cohort included women who used a bisphosphonate and had at least one available pretreatment bone mineral density (BMD) total hip scan. AFFs were identified and verified by physician review of x-ray images for fractures occurring during the study period with ICD-9 codes for subtrochanteric or femoral shaft fractures. Women were considered to have discontinued bisphosphonates if there was a gap of over 3 months between the last bisphosphonate use and cohort entry anniversaries. Researchers included information on potential confounders of the association between discontinuation time and AFF such as age, race/ethnicity, smoking status, fracture history, duration of bisphosphonate use, discontinuation of glucocorticoid use, and pretreatment total hip T-score.

A second study in the same cohort consistently showed that women with higher pretreatment BMD had a larger risk of AFFs.

Researchers assessed the relationship of bisphosphonate duration to AFF risk before and after treatment, including pretreatment BMD in multivariate Cox models. In a multivariate model without pretreatment BMD, those with longer bisphosphonate duration had higher AFF risk. Compared to those taking the drug for less than 1 year, the relative hazard (RH) for 1-4 years of bisphosphonate use was 3 (95% CI, 1.4-7.3), for 4-8 years of bisphosphonate use was 15 (95% CI, 7-33), and for over 8 years was 37 (95% CI, 16-83).

Pretreatment BMD, when added to the model, did not attenuate the relationship of bisphosphonate duration and AFF risk. However, it showed those with higher BMD had a 40% increase in AFF risk per standard deviation of BMD increase (HR = 1.4; 95% CI, 1.2-1.7).

In those with normal pretreatment BMD (T-score greater than –1), the RH for 4-8 years of bisphosphonate use (versus less than 1 year) was 35, compared with 15 in those with osteopenic BMD (T-score –1 to –2.5) and 6 in those with osteoporotic BMD (T-score less than –2.5).

If confirmed in other studies, the results suggest that pretreatment BMD could impact clinical decisions around patient selection for bisphosphonate initiation and drug holidays, Dr. Adams said.

She reported receiving grant and research support from Merck.

SOURCES: Adams A et al. ASBMR 2018 Abstract 1005, and Black D et al. ASBMR 2018 Abstract 1007

Next Article: