site map


Privacy policy  ▪  About


Bookmark and Share

BLOG: March 2010

 Do bone drugs work?

This month, many causal readers could caught headlines reporting of a recent study "adding more evidence that bone drugs work". Articles do not elaborate on what the old evidence is, but they are quick to add that the new study not only confirms that bone drugs work, but that it is also "reassuring" of their safety.

Is it really all that rosy? Let's take a closer look.

The study in question (Bisphosphonates and Fractures of the Subtrochanteric or Diaphyseal Femur, Black et al, 2010) actually focus on a single, rather rare adverse effect associated with bisphosphonate use: specifically, with use of alendronate (Fosamax) and zoledronic acid (Reclast), two widely used drugs for prevention and treatment of osteoporosis.

The study analyzed data from three large bisphosphonate trials (FIT, completed in 1998, FLEX in 2006 and HORIZON-PFT in 2007). In all, there was 14,195 women participants. In the total of 284 fractures (2% incidence rate), there was 12 fractures in ten patients of this particular type (4.2% of all fractures, less than 0.1% incidence rate). The rate in treatment vs. placebo group was nearly identical in the FIT study (1.03, or 3% higher), while somewhat elevated in the other two trials (1.33 and 1.5 in FLEX and HORIZON-PFT, respectively).

Considering very low incidence, the difference is both, statistically unreliable and insignificant. Hence the study concludes that there is no significant increase in risk from this particular fracture associated with bisphosponate use.

Does that makes bisphosphonates - type of synthetic compounds to which bone drugs like Actonel, Fosamax and Zometa/Reclast belong - safe? The very concept of the study, focusing at the very rare adverse effect possibly associated with use of bisphosphonate drugs, may create such impression. But the reality is different.

Bone drugs commonly cause major metabolic imbalances: lower blood calcium, elevated parathyroid hormone (PTH) and phosphate levels Normally, elevated PTH stimulates calcium release from the bone, in order to maintain its needed blood level, but bisphosponates interfere with calcium release. They also interfere with metabolism of other minerals involved in the natural bone formation/resorption cycle.

Knowing how important minerals are for the proper body function, it is no wonder that bisphosphonates are

notorious for quickly making life miserable for many of its users.

Of course, you won't find that in your doctor's manual, effectively written by the drug manufacturer (part of it is that users often are not aware that their problems are drug-related, so it goes unreported). A quick look into uncensored online forums is recommended, if you want to get the real taste of it.

Among common adverse effects of bone drugs are fever, bone, joint and/or muscle pain, upper GI tract irritation and esophageal ulceration/erosion (possibly leading into esophageal cancer). Among less common but possible, are atrial fibrilation (abnormal heart rhythm), nephrotic syndrome (kidney degeneration) and jaw osteonecrosis.

Can we expect better with drugs that nearly stop bone resorption, causing almost as much reduced bone formation? They literally lock minerals inside the bone, nearly halting the natural process of bone regeneration. But this is not making the body - which normally has control over the mineral flow, and for a good reason - happy.

How does a bone drug lock minerals in the bone? Your doctor can't tell you, because drug manufacturer didn't tell him. No one knows the exact mechanism. But we know enough about bisphosphonates and body metabolism to follow some likely trails.

Bisphosphonates are closely related to pyrophosphate (PPi), one of the key metabolic intermediaries. PPi is a natural inhibitor of mineralization (i.e. mineral flow/exchange) in the blood and urine, which cannot get into the bone because it gets destroyed by alkaline phosphatase in the bone lining. Obviously, the body

doesn't want a mineralization inhibitor in there.

But bisphosponates, thanks to the twist in their molecular structure, make it into the bone, and take hold of the minerals.

That makes the bone more dense, but does it also make it stronger? There are solid indications that the gain in bone strength - measured by the frequency of fractures - does not follow this artificially created density, and may be marginal at best.

For instance, among 1609 postmenopausal women in the EPIC study (Ravn et al, 1999), divided in two treatment groups (Fosamax or estrogen-progestin) and placebo, there was no significant difference between the groups in fracture incidence over 4-year period. This is certainly influenced by relatively small number of participants (and fractures), but the absence of significant effect in three groups of 500+ women over 4-years period

speaks clearly about the actual efficacy level of these treatments.

Likewise, among 1099 postmenopausal women in the FLEX trial (Black et al, 2006), which was looking to establish optimum duration of treatment with bone drugs (Fosamax), the half that after 5 years of drug treatment was put on placebo for another 5 years, had essentially identical number of fractures as the half that continued taking the drug, despite losing more than half of their bone density gain.

There was no difference in severity of fractures either, but the extended-treatment group had significantly greater height loss: 3.5cm vs. 2.1cm in the placebo group. This doesn't add up with the denser, presumably stronger bones, does it?

Of course, some studies did come up with possibly significant gains fractures-wise from taking bone drugs. Afterall, there would be no market for them without it. The prime example, the Fracture Intervention Trial (FIT, Black et al, 1998), did not find significant benefit of Fosamax in reducing clinical fractures overall, but did find that it reduced clinical fractures in women with low bone density (i.e. established osteoporosis) by 36%. It also reduced the overall rate of radiographic vertebral fractures by 44%.

Those were the numbers that sold the drug. But how reliable are they? Neither clinical examination nor visual radiographic assessment are very accurate diagnostic methods. In the mentioned FLEX study, placebo group had more than double the incidence of "clinically recognized vertebral fractures" vs. treatment group (5.3% vs. 2.4%, respectively). However, with more accurate quantitative morphometry it reduced to the insignificant 15% higher incidence (11.3% vs. 9.8%).

Also, it is very obvious that the study numbers

are not representative of this sub-population group (54-81y),

even without going into its methodology shortcomings. Participants for this study are picked trough the careful selection process: out of 54,000 women that responded to over 1 million mailings, only 4,432 are admitted. Less than 1/5 of those that were not had too high bone density (BMD over 0.68 g/cm3); the rest of 40,000 were ineligible on some other basis, or "did not wish to continue".

Among exclusion criteria were ulcers, indigestion, kidney or liver condition, high blood pressure, unstable angina, hyperthyroidism, hypothyroidism and hyperparathyroidism. In all, less than 1 in 13 of the women interested in participation were admitted to the study. The selection narrowed participants very much to those

least likely not to tolerate, or respond to the treatment.

Considering vested interest of the study sponsor - Merck, the drug manufacturer - this bias, and possibly some others, were part of the strategy geared toward tipping the scale toward producing desired result.

Despite hand-picking participants, the study showed significant benefit of taking Fosamax only for women with established osteoporosis, but even that, as noted,  was based on unreliable diagnostic methodology.

Also, nearly 10% of the woman on Fosamax did not gain in bone density; there is no study figure for the proportion of woman for which the gain was significant, but it is certainly less than 90%. In a real-life sample of woman, this inherent inefficiency rate of bone drugs is likely to be considerable.

By the way, the other major study concerning bone drugs, the above mentioned FLEX (also sponsored by Merck), used only women that participated in the FIT trial, but with additional requirements. In order for the FIT women to be admitted to FLEX, they had to:

have better BMD (bone mineral density) than they had
at the beginning of the FIT trial, and

have hip BMD above determined minimum level (better than
3.5 standard deviations below the average for young adult woman)

In other words, women carefully selected to fit the previous manufacturer's study were subjected to another filter, excluding those that had no positive response to the drug, and those most likely to suffer fracture.

Now, anyone with the money to spend can conduct a medical study in any way they want, but studies with so clearly biased selection process are not representative of the given segment of general population,


Did not bother the government a bit. Nor that the studies flirted with half the dose recommended for treatment in practice guidelines, in order to minimize adverse side effects. Large bisphosphonates trials have shown that lower doses do not have significantly different fracture rates than higher doses, and can even be more effective.

For instance, a group of 5445 women 70-79y old in risedronate (Actonel) trial (McClung et al, 2001) had statistically significant relative risk reduction with the 2.5mg/d dose (confidence interval 0.3-0.9), but not with 5mg/d dose (CI 0.4-1.1). Despite that, manufacturer (Warner Chilcott, Sanofi-Aventis) went with the higher dose as a treatment recommendation. And so did Merck with Fosamax (5mg/d vs. 10mg/d).

Makes more money. More side effects seems to be of secondary importance.

Likewise, as the results of the FLEX study showed, there is no benefit of reduced fractures from the treatment extended beyond 5-year period. There are even indications that prolonged treatment can be counterproductive. Analysis of the database of the Dutch PHARMO trial, including 14,750 women in the 1996-2004 period, showed that women who took bisphosphonates for 3-4 years had 39% lower fracture incidence than those who took it only for 1-2 years, but those who took bone drugs for 5-6 years

had 12% more fractures than those taking them
for less than a year.

Study conclusion is that there is a positive correlation between bisphosponates use and reduced fracture risk, but that such link is "inconclusive" (read: non-existent according to the data) for treatment duration of over 4 years (Meijer et al, 2008).

This is, in fact, consistent with the general scheme of how bone drugs work, shown below.

Most of the increase in bone density due to bone drugs (when they work) is due to disparity between the rates of shutting down bone resorption and bone formation. The former is very quickly reduced from about 7-8% average for post-menopausal women to below 1% (as a portion of mineralizing vs. total surface of the trabecular bone), while bone formation continues at decreasing rate up to a year, or so. The net result is increased bone mass. About 2-3 years into treatment, bone density increase reaches its plateau, after which begins its gradual decline.

Prolonged use of bisphosphonates causes structural bone damage (accumulation of micro-cracks) in experimental animals, and investigation showed similar effect in humans (Dobnig et al, 2006). In another recent study (Solomon et al, 2008), post-fracture bisphosponates use was associated with 2.4 times higher risk of non-union.

Nevertheless, there is no hint of limiting treatment duration with bone drugs in the practice guidelines.

A little can of worms. Anyway, after taking a little closer look, bone drugs don't seem to be so reassuringly effective and safe. Effectiveness-wise, the risk of fracture even in the high-risk subpopulation of post-menopausal woman is relatively low to begin with. Reducing it by 30-50%, as most studies indicate, does not change much in the overall picture. For illustration, a 40% risk reduction would change 1 in 20 (yearly) risk to 1 in 33. From 5 to 3 women in every 100 women. In other words,

only about 1 in every 50 high-risk women taking bone drug
actually benefits from it.

For the medium-to-low risk subpopulation, taking bone drug benefits only one in up to a few hundred women. That is based on results within selected study populations. In the real life setting, the effectiveness of bone drugs is likely significantly lower.

On the other hand, their adverse side effects are real, and likely to be significantly more of a problem than what the study-setting results indicate. Simply put, nearly shutting down such a vital part of the body function as it is bone regeneration invites trouble. Those not wishing to test their luck will find quality nutrition, regular physical activity and, if necessary, a simple devices like vibrating platform, both more effective and safer way of preventing/strengthening weak bones.