| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
| ||||||||||||||
BOOK REVIEW |
Assistant Professor of Medicine Division of Geriatric Medicine Dalhousie University Halifax, Nova Scotia, Canada B3H 2E1
Osteoporosis (2nd ed.), edited by Robert Marcus, David Feldman, and Jennifer Kelsey. Academic Press, San Diego, CA, 2001, 2 volumes, 1672 pp., $399.99 (cloth).
Osteoporosis and the fractures that result from this disease were once viewed as an inevitable consequence of aging, for which no effective therapy could be offered. Over the past few decades, there has been an incredible growth in both our understanding of the basic biology of bone, and in the number of evidence-based therapies that have become available to prevent fractures. This body of literature has been comprehensively reviewed in Osteoporosis, edited by Robert Marcus, David Feldman, and Jennifer Kelsey. Authored by experts from a number of different countries, it covers topics ranging from normal bone physiology, through epidemiology and risk factors, to pathophysiology, evaluation, and management.
Many therapies have now been proven to treat osteoporosis. These include calcium and vitamin D, bisphosphonates, estrogen and selective estrogen receptor modulators (SERMs), calcitonin, and parathyroid hormone. In this book, discussion of each of these therapies is divided between at least two chapters, one that focuses on basic biology and one that is clinically focused. Although this strategy inevitably results in duplication of material, the book is well edited and the chapters are clearly cross-referenced.
Calcium and Vitamin D
In elderly patients, there is much evidence indicating that calcium and vitamin D deficiencies contribute to osteoporosis. This is discussed in Chapters 9, 40, 67, and 68. The average North American diet today falls far below the recommended intake of these nutrients. Chapter 27 provides interesting support for the idea that human physiology evolved to protect us from calcium toxicity, but in modern times, these protective mechanisms may be detrimental. As Robert P. Heaney reports:
Calcium and vitamin D were present in superabundance in the environment in which the human species evolved.... In equatorial east Africa, with ample sunlight year round, mechanisms coevolved that prevented accumulating an excess of vitamin D.... As a result, vitamin D accumulation in the skin plateaus after a few minutes of sun exposure... In latitudes such as that of Boston and farther north, the sun is so low in the sky in winter that effectively none of the responsible ultraviolet rays gets through the atmosphere, even on a sunny day.... The plant foods eaten by hunter gatherers provided a calcium intake that...would have been in the range of 2000–4000 mg/day....The median value...in the United States [is] 600 mg/day. Sources...included greens, tubers, roots, nuts and berries.... Invertebrate and reptilian sources of animal protein typically have calcium-to-calorie ratios six-fold higher than fish or mammalian meats. In contrast, cultivated cereal plants, legumes and fruits—the plant foods that modern humans mainly consume—exhibit augmented levels of carbohydrate and/or fat without a proportionate increase in minerals and vitamins; thus they almost always have lower calcium densities.... Because...early human diets were very rich in calcium, the human intestine either failed to develop effective transport mechanisms or actually developed an absorptive barrier to protect against too much calcium. Nor did mechanisms to conserve absorbed calcium develop. (vol. 1, pp. 675–676)
Surveys demonstrate a high prevalence of vitamin D deficiency in elderly persons. Many authorities also argue that the lower limit of normal for 25-hydroxy vitamin D is set too low. Adding to a poor diet and low sunlight exposure is the high prevalence of renal dysfunction, which both impairs conversion of 25-hydroxy vitamin D to the active 1,25-dihydroxy vitamin D, and also leads to increased phosphate levels. Elderly persons may also have decreased absorption of calcium from the gut. All of these factors are believed to increase levels of parathyroid hormone, which may lead to increased bone resorption. A study that randomized institutionalized women in France to calcium and vitamin D supplementation or placebo demonstrated a significant 43% reduction in hip fracture at 18 months in the active treatment group. This benefit persisted (29%) at 3 years. The difference was also significant when calculated as intention-to-treat. Community-dwelling elderly men and women in the United States, when receiving supplements of calcium and vitamin D, had a significant 50% reduction in nonvertebral fractures. Studies with vitamin D alone have had mixed results. However, a recently published British study that randomized elderly community-dwelling persons, mostly men, to vitamin D 100,000 IU every 4 months for 5 years did show a significant reduction in all fractures, as well as the total number of hip, wrist or forearm, and vertebral fractures (Trivedi, Doll, & Khaw, 2003). The recommended daily intake for calcium and vitamin D has risen in recent years to 1200–1500 mg of calcium and 800–1000 IU of vitamin D. This cheap and innocuous therapy has the potential benefit to dramatically reduce the number of fractures, and thereby reduce morbidity and mortality.
Pharmacological Therapy
Chapters 69, 70, 72, and 73 provide a thorough review of the evidence for the bisphosphonates, estrogen, raloxifene, and calcitonin. All of these medications work by decreasing bone resorption. The fracture reduction is reduced more than would be predicted by the gain in bone mineral density, leading to the hypotheses that these agents work by decreasing the depth of the resorption pits and/or by increasing the time available for mineralization by slowing bone turnover.
The best-proven agents are the bisphosphonates, alendronate and risedronate. Several large randomized trials with alendronate that enrolled women with osteoporosis have shown approximately a 50% reduction in radiographic vertebral fractures, clinical vertebral fractures, forearm fractures, and hip fractures. Risedronate trials have shown a 40–50% reduction in radiographic vertebral fractures, a 33–39% reduction in nonvertebral fractures, and a 40% reduction in hip fracture. The most common side effects from these medication are gastrointestinal, the most serious of which is esophageal ulceration. Compliance is an issue for many patients, especially frail elders and those with cognitive impairment. The drugs have poor oral absorption, and must be taken on an empty stomach with an 8-ounce glass of water. The person must then remain upright and wait 30 minutes before eating. The availability of once-weekly dosing has partially overcome this problem.
Estrogen therapy is reviewed in Chapter 69. At the time this book was written, the only evidence for nonvertebral fracture prevention was epidemiological studies, and the cardiovascular benefits of estrogen were uncertain. Since that time, an article has been published by the Writing Group for the Women's Health Initiative Investigators (2002) that showed a 34% reduction in hip fracture and a 24% reduction in combined fractures among patients on estrogen and progestin. However, there was also an increased risk for coronary artery disease, stroke, breast cancer, and venous thromboembolism. Raloxifene, reviewed in Chapter 70, is a selective estrogen receptor modulator. The Multiple Outcomes of Raloxifene Evaluation (MORE) trial showed a reduced risk of vertebral fracture, but there was no decrease in nonvertebral fractures. Evidence about cardiovascular risks or benefits awaits the results of an ongoing trial. As reviewed in Chapter 73, a recent study with calcitonin led to difficult-to-interpret results with a significant reduction in vertebral fractures in the 200 IU dose of nasal calcitonin, but no change in the 100 IU or 400 IU dose. This study also had a large (59%) dropout rate.
Parathyroid hormone, discussed in Chapters 7, 40, 49, and 77, is an exciting addition to the current medications available to treat osteoporosis. In contrast to the other available therapies, which work by decreasing bone resorption, parathyroid hormone, when given intermittently, is an anabolic agent that increases bone turnover and results in new bone formation. Once daily subcutaneous injections with the 1-34 amino fragment have been demonstrated in a clinical trial to decrease vertebral fractures by 65% and nonvertebral fragility fractures by 50%. Barriers to the use of parathyroid hormone include the cost, the need for subcutaneous injections, and the fact that it has not yet been proven to reduce hip fractures.
Decreasing Falls
In most circumstances, fractures are the result of a fall. Thus, as discussed in Chapter 32, interventions must be targeted at decreasing falls, and not focus solely on pharmacological measures to increase bone mineral density. As the cause of falls in an elderly person is often multifactorial, multifaceted intervention strategies have been developed that focus on such things as strengthening and balance exercises, decreasing medications and reducing home hazards. These studies have shown a reduction in falls, but have not been powered to show a reduction in fractures. An individualized in-home exercise program targeted at women older than age 80 did show a reduction in falls with injury (Campbell et al., 1997). Medications such as benzodiazepines, antidepressants, and neuroleptics have been correlated with falls in many studies. A randomized trial has demonstrated a reduction in falls when subjects were tapered off psychotropic medications, but long-term compliance was poor. Hip protectors may also reduce the risk of hip fracture, but again compliance has been a challenge.
Underdiagnosis
Unfortunately, in spite of the availability of a number of therapeutic options, persons with osteoporosis remain underdiagnosed and undertreated. The exact prevalence of osteoporosis is difficult to determine and, of course, depends on how osteoporosis is defined. The definition referred to in Chapter 35 is "a disease characterized by low bone mass and microarchitectural deterioration of bone tissue, leading to enhanced bone fragility and a consequent increase in fracture risk" (vol. 2, p. 5). This defines low bone strength as the disease, rather than the fracture. Bone strength consists of both bone mass or density and bone quality, but as bone quality is difficult to determine clinically, most research has focused on bone density. The World Health Organization defines osteoporosis as bone mineral density (BMD) lower than 2.5 SD below the mean for a normal young adult. As discussed in Chapter 21, the prevalence of osteoporosis in postmenopausal women, as assessed by dual energy X-ray absorpitometry (DEXA) at the femoral neck, is 20%. In women older than 80, the prevalence is 50%. If defined as BMD lower than -2.5 SD at either the lumbar spine, hip, or mid-radius, the prevalence is approximately 20% in White women 60–69 years of age and rises to 70% in those older than 80.
The prevalence of fragility fractures is also high and increases with age. The lifetime risk of one of the three traditional osteoporotic fractures (hip, distal forearm, or vertebral) is estimated to be about 40% for a 50-year-old White woman. In American women older than 65 years of age, the prevalence of hip fracture is 5% and that of vertebral fracture is 12–24%. However, these numbers are clearly an underestimate of the total number of osteoporotic fractures, as proximal humeral, pelvic, proximal tibial, and distal femoral fractures are common with increasing age, and are also associated with low BMD.
Screening
As discussed in Chapters 34 and 61, osteoporotic fractures frequently result in devastating consequences, including acute and chronic pain, loss of function and independence, institutionalization, increased mortality, and, of course, increased health care costs. A survey of women older than age 75, conducted by Salkeld and colleagues (2000), revealed that 80% of women would rather be dead than suffer a hip fracture that led to nursing home placement. In the same way that one would argue for identifying and treating hypertension before the stroke occurs, the argument has been made for screening to detect low BMD before the fracture occurs.
However, evidence to support pharmacological treatment of patients with a low BMD who have not yet had a fracture is limited, as most of the subjects enrolled in clinical trials had preexisting vertebral fractures. The Fracture Intervention Trial-2 (FIT-2; Chapter 72) studied the effect of alendronate in a patient group with low BMD, but without fractures. The subgroup with a BMD lower than -2.5 SD did have a reduction in radiographic vertebral fractures and a significant decrease in all clinical and hip fractures. In the MORE trial (Chapter 70) of raloxifene, persons with a BMD below -2.5 SD without a preexisting vertebral fracture had a 50% reduction in vertebral fracture. But, as discussed earlier, there was no reduction in nonvertebral fractures in the MORE trial. The Women's Health Initiative study (Writing Group for the Women's Health Initiative Investigators, 2002) showed a clinical fracture reduction with the use of estrogen and progestin. No information was provided on BMD or pre-existing fractures. In the risedronate hip fracture study (Chapter 72), subjects with a BMD below -2.5 SD, who did not have a preexisting vertebral fracture, showed a 40% reduction in hip fracture, but it was not statistically significant.
Although women with the lowest BMD have the highest individual risk of sustaining a fracture, a large proportion of the fractures occur in the bigger subgroup with BMD above -2.5 SD. Johnell and colleagues (2002) reported a post-hoc analysis of the MORE trial that divided women without preexisting vertebral fracture into BMD subgroups. Osteopenic women, with BMD scores at the hip above -2.5 SD, were reported to have a reduction in radiological and clinical vertebral fractures. This study has been published in abstract form only. In the FIT-2 trial (Cummings et al., 1996), there was no significant reduction in vertebral or clinical fractures for women with a bone density between -2.5 and -2.0 SD or those between -2.0 and -1.6 SD. In fact, in the latter two subgroups there was not even a trend for reduction of clinical fracture, as the relative hazards for clinical fractures were 1.03 and 1.14, respectively.
Given this evidence, the question arises in Chapters 23, 33, and 57 as to who should be screened to detect a low BMD. This comes down to a judgment as to a reasonable number needed to screen to detect a person with a BMD below -2.5 SD, and a reasonable number needed to treat to prevent a fracture. For example, only 4% of women aged 50–60 have a BMD at the hip below -2.5 SD compared with 30% of women aged 75–80 and almost 50% of those older than age 80. The risk of a fracture is also lower in a younger woman than an older woman, even with the same BMD. For example, a 60-year-old with a BMD of -3.0 SD has a 5-year risk of hip fracture of 3%; an 80-year-old with the same BMD has a risk of 9% (Nelson, Helfand, Woolf, & Allan, 2002). The 10-year risk for all osteoporotic fractures in a 55-year-old with a BMD below -2.5 SD is 17%; for an 80-year-old it is 33% (Brown & Josse, 2002). To prevent one hip fracture in a woman aged 55–59 years, it has been estimated that 4,338 women would need to be screened and 193 treated for 5 years. In women aged 75–79 years, 143 would need to be screened and 41 treated. Preventing a vertebral fracture would require screening 1,338 55–59-year-olds and treating 60, as compared with screening 75 75–79-year-olds and treating 21 (Nelson et al., 2002).
Recent consensus guidelines recommend screening based on age alone only for women 65 years or older (Brown & Josse, 2002; Nelson et al., 2002). For younger women, consensus guidelines vary. The U.S. Preventative Services Task Force guidelines (Nelson et al., 2002) limit their recommendations to those women between 60 and 64 years of age who have low body weight or are not using estrogen. The Canadian guidelines (Brown & Josse, 2002) recommend including postmenopausal women between 50 and 65 years of age who have a positive family history of an osteoporotic fracture, chronic use of steroids, malabsorption, hyperparathyroidism, propensity to fall, osteopenia on X-ray, early menopause, or two of the risk factors of rheumatoid arthritis, past history of hyperthyroidism, chronic anticonvulsant therapy, low dietary calcium, smoker, excessive alcohol, excessive caffeine, low body weight, weight loss, and chronic heparin therapy. Some of these risk factors aim to target testing to those most likely to have a low BMD, although studies have not shown risk factors to be very specific. Other risk factors aim to identify women at a higher risk for fracture, so as to decrease the number needed to treat to prevent a fracture. Although this seems reasonable, clinical studies have not examined risk factors beyond age, fracture history, and BMD.
For men, the issue is even more controversial. So far, screening men for osteoporosis is only recommended in the Canadian guidelines. Fewer men have been studied epidemiologically and fewer still have been included in clinical trials. There is no consensus about what BMD cutoff should be used to define osteoporosis in men. Men have larger bones than women. As BMD is calculated over a two-dimensional area, the volumetric BMD is underestimated in persons with larger bones. This would argue for the use of a sex-specific t-score. However, as larger bones provide a mechanical advantage against fracture, others argue that a cutoff identical to women is more appropriate. Men have a lower incidence of fracture, but a higher morbidity once a fracture occurs.
The advantages and disadvantages of the different radiological screening modalities that can be used to detect low BMD are described in Chapters 33 and 59. Most evidence exists for axial DEXA. However, in many areas, access to these machines is limited. Other modalities, such as ultrasound, are cheaper and portable. There is evidence that these machines do predict future risk of fracture. However, the same individual can have drastically different t-scores on different machines. This is because of biological variability between different anatomical sites, and also because the t-scores for different machines are derived from different young normal populations. Machines other than DEXA have not been used to define subjects enrolled in clinical trials, so cutoffs for treatment have not been determined. Also, ultrasound machines are not useful for following response to therapy.
There has not yet been a study that has evaluated a screening program. One might hypothesize that it would be the younger patients who would be more likely to be screened, rather than frail elderly persons, with multiple medical problems, who are actually more likely to have a low BMD and are at a higher risk for fracture. If 50–70% of women older than age 80 have a BMD below -2.5 SD, should 50–70% of women older than age 80 be treated with a bisphosphonate? That is an astounding figure, but not so different from the prevalence rates for hypertension. There is always a real concern about polypharmacy in these patients, as many are on multiple medications. Some would argue that if a woman has not had a fracture by the age of 80, that is indicative of better quality bones, and perhaps a reduced risk of fracture, but that is only conjecture. Chandler and colleagues (2000) demonstrated bone density to be a predictor of fracture, even among elderly nursing home patients. FIT-2 (Cummings et al., 1996), the trial that demonstrated a benefit for treatment based on low BMD alone, enrolled patients with an average age of 68, but almost 600 patients (13%) were 75 and 80 years of age.
Undertreatment
The group for which there is the strongest consensus is persons with a preexisting fragility fracture. A previous fracture is a strong predictor of a future fracture, independent of BMD. This is likely because a previous fracture is a marker of bone quality that is not captured by BMD. It may also be a marker for propensity to fall, although studies have attempted to control for this. As discussed previously, most of the evidence supporting treatment for osteoporosis has come from studies that enrolled patients with a preexisting osteoporotic fracture. Because persons with a previous fracture have a higher absolute rate of future fracture than persons with an equally low BMD but no prior fracture, they have an absolute greater risk reduction with treatment. Thus fewer persons would need to be treated to achieve the goal of preventing a fracture. One important question is whether bisphosphonates work equally well in persons who are in their eighties or nineties, who are frail, and/or who have more advanced disease. Some physicians believe that it is too late to treat these patients and/or that their bones would be less responsive to therapy. Applying clinical trial evidence to frail elderly patients is difficult in most clinical scenarios because the vast majority of patients enrolled in clinical trials are younger and healthier. However, in the bisphosphonate trials, although the average age was 70 or younger, patients were enrolled up to the age of 80 or 85 years. Subgroup analyses (Ensrud et al., 1997; Watts et al., 2003) that compared age (less than or greater than 75 years for alendronate, 70 years for risedronate), bone density, and number of preexisting fractures showed equal relative risk reduction and greater absolute risk reduction for persons with more severe osteoporosis and for elderly persons. The benefit from these medications is realized in a relatively short time period of 12–18 months.
In spite of the evidence for effective therapies, numerous studies document undertreatment of persons who have had a fragility fracture. This care gap was one subject that was not well addressed in Osteoporosis. Part of undertreatment may be due to underdiagnosis. Many vertebral fractures are silent and do not come to the attention of a physician. Vertebral fractures on X-rays are underreported by radiologists. Should all patients above a certain age or who are osteopenic have a spinal X-ray? The Canadian guidelines address this issue by recommending an X-ray for any person with kyphosis, a height loss greater than 4 cm compared to self-report, or a documented height loss of 2 cm.
Underdiagnosis does not fully explain the problem, however, as many persons with a clearly documented clinical fracture are also undertreated. In a Canadian study (Hajcscar, Hawker, & Bogoch, 2000), patients who were seen in a fracture clinic with fragility-type fractures were followed up by telephone one year later. Less than 20% had been given the diagnosis of osteoporosis (half before and half after that fracture). Only 32% had been prescribed calcium, 13% vitamin D, and 7% a bisphosphonate. Sixteen percent were on hormone replacement therapy. In an American study, Freedman, Kaplan, Bilker, Strom, and Lowe (2000) found similar results among patients who had sustained a distal radial fracture. Only 23% had been started on an antiresorptive prescription medication. They also found that the older the patient, the less likely he or she was to be treated, in spite of the fact that older patients are at a higher risk of having another fragility fracture. Follin, Black, and McDermott (2003) reported the treatment of patients who had been admitted to hospital with a hip fracture. The diagnosis of osteoporosis had been noted in the chart of 14% of patients at the time of discharge and in only 26% over the first year of follow-up. On admission 10% of patients were taking calcium and 7% were taking vitamin D. This increased to only 17% and 14% at discharge and was unchanged over the following year. Use of alendronate was 1% on admission, 2% at discharge, and 6% over the next year. The figures for estrogen were 8%, 9%, and 10%, and for calcitonin were 1%, 1%, and 5%. Overall, at discharge and during one-year follow-up, three quarters of patients were receiving neither calcium, vitamin D, nor any other pharmacological therapy for osteoporosis. Fall risk assessment or education was documented in only 3% of charts. A study in long-term care facilities in Canada documented that only 26% of patients met the recommended 1200 mg intake of calcium and 30% met the recommended intake of 600 IU of vitamin D (Lee, Drake, & Kendler, 2002).
Addressing Undertreatment
How can this care gap for osteoporosis be addressed? There are many barriers. Orthopedic clinics are busy and many surgeons do not see treatment of osteoporosis as their responsibility. The focus is too often on the fracture and not on the underlying cause of the fracture. Polypharmacy, drug side effects, and cost are concerns expressed by physicians. Physician education is definitely an issue. Simonelli, Killeen, Mehle, and Swanson (2002) surveyed primary care physicians and orthopedic surgeons. Ninety-one percent and 100%, respectively, said they would be more likely to treat "if they had a safe medication shown to reduce patients' risk of recurrent fracture by 50% in less than 2 years." This is in spite of the evidence from trials of calcium and vitamin D and bisphosphonates that demonstrate close to a 50% risk reduction in 12–18 months.
The question of how best to address this problem of undertreatment has not been answered. Should patients themselves be targeted through the media? Should calcium and vitamin D be considered a dietary supplement and routinely given to all nursing home patients without the need for a physician's order? And how can this be implemented? Would a mention of the word "osteoporosis" in orthopedic clinic letters change the awareness and behavior of primary care physicians? Should nurse practitioners be involved with patients who have been admitted with a fracture or who are seen in orthopedic clinics? Majumdar and colleagues (2002) conducted a study among patients who were discharged from the emergency room after a wrist fracture. The intervention consisted of both patient counseling, with written materials and phone calls, and individualized written recommendations for the primary care physician for BMD testing and therapy. This resulted in a significant increase in BMD testing in the intervention group (52% vs. 8%) and an increase in treatment with a bisphosphonate (27% vs. 6%). There is also the important nonpharmacological component of the treatment of patients who have fallen. At present, there is a shortage of geriatricians and multidisciplinary teams to implement the multifaceted strategies that have been proven to reduce falls. The benefit of an in-home exercise program has been proven, but in most areas there is no access to trained persons to develop and supervise these exercise programs. Coordination through a home care program may be one option, but this will require extra resources.
In summary, there will undoubtedly be many important discoveries in the future, but we have learned much about osteoporosis over the past few years. This research is documented and discussed in detail in this two-volume book, Osteoporosis, and it would be a helpful reference for any scientist or clinician with an interest in the field. Far from osteoporosis being an untreatable consequence of old age, there is much that can be done to reduce the risk of fracture. The critical question right now is how we can ensure that this knowledge actually translates into a change in clinical practice.
References
| ||||||||||||||
| HOME | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|---|
