Study Finds BMD Can Stand in for Fractures. Will the FDA Agree?

Marlene Busko

October 03, 2018

MONTREAL, QUEBEC — Improvements in bone mineral density (BMD) after 2 years of osteoporosis drug therapy could serve as a surrogate biomarker for decreased fracture risk, a new multidrug meta-analysis suggests.

These findings from the Foundation of the National Institutes of Health (FNIH) Bone Quality Project are being submitted to the US Food and Drug Administration (FDA), said Dennis M. Black, PhD, University of California, San Francisco, here at the American Society for Bone and Mineral Research (ASBMR) 2018 Annual Meeting

"In this large pooled analysis of 22 trials," he reported, "we found that 24-month percentage change in dual X-ray absorptiometry [DXA] BMD explained a large proportion of the fracture risk reduction across a range of [different classes of] osteoporosis treatments."

If the FDA accepts improvements in DXA-BMD as a primary efficacy endpoint in trials of osteoporosis drugs, Black said, it "could very substantially reduce the costs and the time required for new drug development."

"It could really open up the field to a lot of new treatments that would be very helpful, I think we all agree."

To Medscape Medical News, Black elaborated that "the longest, largest trial that's been done in the [osteoporosis-therapy] field was [more than] 15,000 patients followed for 5 years."

In contrast, "if the FDA were to say that you need to show that over 2 years bone density at the total hip region [femoral neck] is increased by about 4% or more, that would be a trial that you could do with maybe...400 or 500 patients" in 2 years.

Invited to comment, John Eisman, MBBS, PhD, Garvan Institute of Medical Research, University of New South Wales, Sydney, Australia, told Medscape Medical News that "what's fascinating about this data is that...all of these [osteoporosis] agents, irrespective of mechanism...seem to indicate the same thing" with respect to the correlation between increase in BMD and decrease in fractures. 

If the researchers could calculate not only relative risk hazard ratios but also absolute risk reduction for fractures then clinicians may be able to say to an individual patient, "this is your risk, and this is what your benefit might be," based on the study data, he added.

FDA, Patient, and Physician Concerns About Safety

In 2013, the FNIH announced the start of a 3-year study to track the progression of osteoporosis using different biomarkers to pave the way for easier approval of new, more effective treatments.

When it launched the Bone Quality Project, the FNIH explained that "recent concerns have prompted regulatory agencies to review the safety of antiresorptive drugs for osteoporosis [bisphosphonates, estrogens, selective estrogen receptor modulators (SERMs), calcitonin, and monoclonal antibodies such as denosumab], while these same concerns among patients and physicians are decreasing the use of this particular class of drugs."

One incident that prompted concern was previously reported in 2014 when, in phase 3 international trials in more than 16,000 women over age 65 years with osteoporosis, investigational oral cathepsin-K inhibitor (odanacatib, Merck), given weekly, reduced fracture rates but led to small increases in serious adverse events including stroke, atrial fibrillation, and atypical fractures.

Two years later, in 2016, Merck stopped the development of the drug after an independent adjudication confirmed the increased stroke rate.

"As you know," Black told the ASBMR audience, "we have this important problem facing the [osteoporosis] field. New therapies are needed but development is challenging."

Current treatments "reduce fractures, but they don't reduce them down to zero," he noted, and there have been safety concerns about things such as atypical femoral fracture, osteonecrosis of the jaw, and rebound vertebral fractures (after stopping therapy).

"They're not perfect," Black summarized, "so we need new treatments."

In this context, the FINH Bone Quality Project was designed to find a surrogate endpoint for bone fractures so that drug approval trials could be done in fewer patients (hundreds instead of thousands) and in a shorter timeframe (2 to 3 years instead of 5 years).

The researchers now have a database with individual patient-level data (including fracture outcomes) from more than 150,000 patients who took part in more than 40 placebo-controlled randomized clinical trials of osteoporosis therapies.  

They plan to investigate three types of potential surrogate biomarkers for fractures:

  • BMD determined by DXA;

  • Quantitative CT-based finite element analysis (QCT/FEA) of bone; and

  • Bone turnover markers measured in blood and urine. 

Increases in BMD Align With Reduction in Fractures 

The current study looked at the first biomarker (DXA-BMD) based on data from 22 trials of osteoporosis therapies in more than 90,000 patients who had at least one determination of DXA-BMD. It was designed to meet new FDA requirements for surrogate biomarkers announced last year.

As part of these requirements, investigators need to perform a meta-regression analysis of relevant published trials and determine the proportion of treatment effect that is explained by the surrogate.

The osteoporosis drugs in this meta-analysis were bisphosphonates (alendronate, ibandronate, risedronate, zoledronic acid), SERMs (arzoxifene, bazedoxifene, lasofoxifene, raloxifene), odanacatib, denosumab, anabolic drugs (teriparatide [parathyroid hormone (PTH) 1-34], abaloparatide [PTH 1-84]), and hormone replacement therapy (including estrogen).

In the regression analysis, the percentage difference in increase in BMD (change in BMD with treatment minus the change in BMD with placebo) at the hip, femoral neck, and lumbar spine was significantly associated with relative risk reduction of having a fracture.   

In a second type of analysis, the researchers determined the proportion of the treatment effect (reduction in fractures) that was explained by changes in BMD

Previous trials of individual osteoporosis drugs have reported that changes in BMD account for 4% to 70% of the treatment effect for different types of fractures.

However, the current study — the first to combine multiple trials and look at fracture outcomes in a uniform way — found that changes in BMD accounted for about 30% to 60% of the decreases in risk of fractures.

Percentage Decreases in Fractures Explained by 2-Year Changes in BMD*

Fracture Type** Hip BMD Change

Femoral Neck
BMD Change

Lumbar Spine
BMD Change
Vertebral, % 60 63 30
Hip, % 50 46 41
Nonvertebral, %  65 70 51
*All P ≤ 0.017.
**4658 vertebral fractures, 894 hip fractures, and 6873 nonvertebral fractures.

"Initial FDA Signals Are Positive"

"We are confident that DXA-BMD can be qualified as a surrogate marker for fracture risk," Black asserted, "because our database is very large, very complete, and robust."

The findings also "compare favorably" with accepted surrogates in trials of drugs in other therapeutic areas (for example, LDL-cholesterol and HbA1c are accepted surrogates for hard outcomes in cardiology and diabetes drug trials, respectively), he noted.

"The initial signals are positive, which means we are still talking about a couple of years," before BMD could potentially be accepted as a surrogate marker for fracture risk in future osteoporosis drug trials, he concluded.  

Black is a consultant for Radius Pharmaceuticals. Asahi-Kasei is on the speakers' bureau for Roche Diagnostics. Eisman has reported no relevant financial relationships.

American Society of Bone and Mineral Research Annual Meeting. September 28 - October 01, 2018. Montreal, Quebec. Abstract 1070. 

For more diabetes and endocrinology news, follow us on Twitter and on Facebook.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.