Are we overtreating osteoporosis?

A patient, keen to make certain that I stay up to date, forwarded to me a copy of a New York Times article from September 8th that suggested doctors are overtreating osteoporosis with drugs that are potentially dangerous.

The very quick and simple answer to that article is a resounding NO! Osteoporosis is under-recognized and undertreated.

The issue that the article addressed is “osteopenia,” a term that is both overused and incorrect.

“enia” is derived from the Greek word for “deficiency” and is appropriately used in many areas of medicine, but bone disease is not one of them. Does osteopenia mean too few bones, akin to neutropenia meaning too few white cells?

The problem with the terminology for bone mineral density measured by DXA is that dividing patients into groups with labels (normal, low bone mass [osteopenia], and osteoporosis) is wrong because BMD is a continuous variable.

Defined cutpoints for intervention are based on extensive epidemiologic research and are subject to ongoing surveillance and change. The cutpoints are guidelines, not speed limits, and must be used in the correct clinical context. Not everyone with hypercholesterolemia will have a heart attack and not everyone with a heart attack has hyperlipidemia.

The same holds true for BMD. While everyone with a T-score of –2.5 or lower has osteoporosis, most fractures related to bone fragility occur in patients with a T-score better than –2.5. Once that patient has sustained a fragility fracture, the diagnosis is osteoporosis, not “osteopenia with fracture.”

The International Society for Clinical Densitometry strongly discourages the use of the term osteopenia, but that is the term that is provided in the printout from most DXA instruments. For many patients, “osteopenia” sounds more ominous than osteoporosis. Almost all of the FDA-approved therapies for the treatment of osteoporosis are also approved for the prevention of osteoporosis, and maybe it is OK to treat “osteopenia” to prevent subsequent osteoporosis.

More often than not, based solely on the T-score diagnostic label, treatment is not indicated. To most of us, and particularly to our patients, a negative number implies loss. As I have mentioned before, it is not possible to make such a dynamic statement, “loss,” on the basis of a single timepoint measurement.

So we have a dilemma! Most fractures occur in patients with a T-score better than –2.5, but most patients with a T-score better than –2.5 do not need therapy. FRAX is a suitable solution, but it may provide misleading information for younger patients. There is much more to be gained from taking a complete history that includes a personal or family history of fracture, diseases and medications that may have adverse effects on skeletal health, and a nutritional history. If that does not clearly indicate an increased potential for bone loss, a standard biochemical and hematologic profile (usually already on file) is appropriate.

There is debate about routinely checking thyroid function in such patients. If you think decision-making about “treating the number” in bone disease is tough, read the literature about treating the number in asymptomatic thyroid disease. Many clinicians routinely measure 25-hydroxyvitamin D, but the cost of that test would cover two years' worth of over-the-counter vitamin D supplement at 1,000 IU per day. There is also no clear indication to measure parathyroid hormone. Again, as I have mentioned in earlier blogs, if none of the above information suggests the need for intervention, the closer the T-score is to –2.5, the more likely I am to order a 24-hour urine collection for measurement of creatinine, calcium and sodium. Hypercalciuria (>4mg/kg body weight/day) likely robs the skeleton of calcium. If the urine sodium is also elevated, a low sodium diet, which is good for anyone, will quickly reduce urine calcium excretion. If the urine calcium is high but the urine sodium normal, hydrochlorothiazide will markedly decrease urine calcium losses. Several epidemiologic studies have demonstrated that hydrochlorothiazide use is associated with a decreased risk of hip fracture.

Still undecided?! Measure fasting levels of a marker of bone resorption, either serum CTX or NTX. These tests cannot be used to diagnose osteoporosis or any other metabolic bone disease, but they will give you an idea about possible bone loss during the next few years. Any value above the reference interval for premenopausal women is suggestive of potential bone loss in the two-year interval before the next DXA and should tilt your thinking toward pharmacologic intervention. If that is the case, you will most likely use one of the antiresorptive drugs approved for the prevention of osteoporosis. You can then repeat the measurement after three months of therapy and have something positive to share with the patient — that the therapy is indeed inhibiting bone resorption.

The concept of bone remodeling (formation and resorption) will be new to most patients, and you don’t have the time needed to explain this properly. Have your patient go to the website of the National Osteoporosis Foundation^* which is very user friendly and has material specifically targeted toward patients.

Yes, we are overtreating BMD numbers in patients who don’t yet need treatment.

Yes, we are undertreating BMD numbers in patients who are at increased risk for ongoing bone loss and do need treatment as soon as you establish that is the case.

And yes, we are most definitely undertreating patients with osteoporosis-related fractures!

This is no different to the approach to any “disease” where the numbers drive our decision-making.

Not quite true! The osteoporosis field is better! No drug for osteoporosis is FDA approved until it has demonstrated, not just that it can affect a BMD number, but that it decreases fracture occurrence in controlled clinical trials.

*Dr Kleerekoper is a member of the Board of Trustees of the National Osteoporosis Foundation.