Osteoporosis and associated fragility fracture is a much more common problem than breast cancer for postmenopausal women. The lifetime risk of a hip fracture is 15%[1] and the risk of any fragility fracture is approximately 50%, while the lifetime risk of breast cancer is 12%.[2] Despite these sobering statistics, osteoporosis is less feared, and most women do not undergo the appropriate guideline-recommended screening for bone health. The evaluation for and treatment of osteoporosis is of particular importance to medical oncologists and other health professionals who treat cancer patients because female cancer patients may receive treatments that increase their lifetime risk of developing osteoporosis and fracture.[3] In addition, the safest options to reduce the risk of osteoporosis in women with a history of breast cancer may be affected by the patient's risk of breast cancer occurrence/recurrence and the type of adjuvant therapy received. It is important that bone mineral density (BMD) be appropriately monitored. Further, in high-risk patients, BMD should be protected by specific treatment, so that women with breast cancer can receive the full benefit of recent advances in cancer therapy. The principles that need to be followed for cancer patients do not differ from those for women within the general population. These include the following:

• Identify by history and physical examination which women are at particular risk for osteoporosis.
• Encourage general health habits that help all women to reduce the risk of osteoporosis.
• Decide who needs the "gold standard" test, ie, the measurement of BMD.
• Decide who might be considered for additional pharmacologic treatment and what treatment is appropriate.
• Reassess risk and, in treated women, reassess treatment efficacy.

Here we provide a general practical overview of these issues generated from guidelines and overviews by the American Association of Clinical Endocrinologists,[4] National Osteoporosis Foundation,[5] Osteoporosis Society of Canada,[6] American Society of Clinical Oncology (ASCO),[7] and other sources. Assessment of Risk by Patient History and Physical Examination Most reviews of osteoporosis define risk factors as being either "major" or "minor." Major risk factors roughly double the risk of osteoporosis and fracture, whereas minor risk factors have a lesser impact. Major risk factors include age > 65 years, history of prior fragility fracture, family history of fragility fractures, and hypogonadism or menopause before age 45. Some guidelines also include being lean as a risk factor of particular importance. Minor risk factors include smoking, history of hyperthyroidism, and low body weight. Using the ASCO guidelines, women at high risk of osteoporosis are identified as:

• All women > age 65 years
• All women aged 60 to 64 years with a family history, body weight < 70 kg, history of nontraumatic fracture, or other risk factors
• Postmenopausal women of any age receiving aromatase inhibitors
• Initially premenopausal women who experience therapy-induced early menopause

Risk factors for osteoporosis evident by history and physical exam are outlined in Table 1. It is immediately apparent that many breast cancer patients might be considered as having risk factors as a result of their therapy, specifically therapy that causes early menopause and/or affects hormone levels. It must be noted that there is little detailed information about the impact of cancer treatment on the risk of bone fracture. Most trials, even though they include women who undergo therapy- induced menopause, neither report problems with osteoporosis nor report on changes in BMD. However, the lack of a demonstrated association is not altogether reassuring, because most trials are underpowered to look for an effect on fractures and do not specifically collect these data. Since the risk of fracture in premenopausal women is low, providing enough statistical power to demonstrate an effect is very difficult to achieve. In addition, the most important effects of fracture risk may be delayed for 10 or more years, and there are almost no data collected on these delayed events. Early Menopause Induced by Therapy
When early menopause occurs after chemotherapy, it is clearly associated with accelerated loss of BMD and should be considered a major risk factor.[8] Early menopause caused by non-chemotherapy-associated ovarian suppression might also be considered a major risk factor. This group of women also experience accelerated loss of BMD. The risk of osteoporosis for women who have reversible menopause due to luteinizing hormone-releasing hormone (LHRH) analogs is somewhat harder to define. Results from the Zoladex Early Breast Cancer Research Association (ZEBRA) trial show that 1 year after treatment with goserelin(Drug information on goserelin) (Zoladex) has been completed, BMD levels are at least partially restored.[9] A cautious view would be to consider patients with early menopause (even if reversible) as having a major risk factor. Adjuvant Treatment With an Aromatase Inhibitor
The best source of data on aromatase inhibitors comes from the Arimidex, Tamoxifen Alone or in Combination (ATAC) trial, involving over 9,000 postmenopausal women with early breast cancer.[10] Although anastrozole(Drug information on anastrozole) (Arimidex) has been shown to be well tolerated with few adverse events, patients receiving anastrozole had a statistically significant increase in the risk of fracture compared with those receiving tamoxifen(Drug information on tamoxifen) (7.1% vs 4.4%, P < .001) at a median treatment duration of 37 months.[11] What must be remembered is that tamoxifen has a modest protective effect against bone fractures (by a relative 20% vs placebo), as demonstrated in the Breast Cancer Prevention Trial (BCPT).[12] Thus, the relative increased risk seen in the anastrozole group in the ATAC trial is likely to be partly due to a protective effect of tamoxifen in the tamoxifen- treated group, and partly to the impact of the lowering of estrogen levels due to anastrozole. These results are consistent with the results of the Canadian MA.17 trial, which randomized patients who completed 5 years of tamoxifen between an extension of hormonal therapy with letrozole(Drug information on letrozole) (Femara) or treatment with placebo. In this trial, there was a small trend toward increased osteoporosis and fracture in the letrozole-treated group,[13] but it should be noted that the comparator group received placebo, not tamoxifen, which would have shown modest protective effects. Additional differences between the ATAC and MA.17 trials include the fact that in the ATAC trial, a prespecified check-list was used by the clinician at every patient visit to collect fracture data, whereas side effects in the MA.17 trial were self-reported. Adjuvant Treatment With Tamoxifen
In the BCPT, a slight non-statistically significant decrease in risk of fracture was seen in tamoxifen-treated premenopausal patients.[12] Thus, the modest increase in the rate of BMD loss in premenopausal patients[14] does not seem to translate into an elevated risk of fracture-at least not in the short term. Treatment of postmenopausal women with adjuvant tamoxifen has modest protective effects on both BMD[15] and shortterm risk of fracture. However, there are concerns about the long-term use of tamoxifen beyond 5 years. Long-term treatment has been associated with a late lack of reduction (perhaps even worsening) in the risk of breast cancer recurrence and continued risk of thrombotic events and endometrial cancer.[12,16-19] Therefore, its use for more than 5 years is generally considered contraindicated.[20] Adjuvant Treatment With Chemotherapy in Postmenopausal Patients
Although small case-matched series have been used to suggest that chemotherapy might have a direct negative effect on BMD,[21] the results of these studies are not convincing. For example, one of the strongest case-matched studies found no statistically significant differences in T-score measurements of BMD between patients who did and did not receive chemotherapy.[22] Who Needs Additional Screening? Deciding Who Should Undergo BMD Measurement
For patients identified by history and physical assessment as being at risk, a measurement of BMD is crucial. Since BMD is the strongest predictor of fracture risk, it is used to define the risk state of patients. The impact of BMD on the risk of fracture is demonstrated in Table 2.[23,24] It shows that age and BMD independently contribute to the risk of fracture. For every fall in BMD T-score by 1, the risk of fracture approximately doubles. There are several ways of screening patients with quantitative measures of BMD. The "gold standard" is dual x-ray absorptiometry (DXA), carried out on the spine and hip. This test can be completed in a few minutes and involves a radiation exposure that is approximately one-tenth that of a chest x-ray. Other methods for screening include peripheral DXA (eg, of the forearm, finger, or heel), quantitative computed tomography, and ultrasound densitometry (eg, of the heel, tibia, or patella). The recommendations of the National Osteoporosis Foundation (Physician's Guide) are typical of guidelines for who should receive a BMD test. This list includes:

• All women > age 65 years, regardless of risk factors
• Younger postmenopausal women with one or more risk factors (other than being white, postmenopausal, and female)
• Postmenopausal women who present with fractures

The recommendations from ASCO outlining who should receive a BMD test are similar and include:

• All women > 65 years old
• Women aged 60 to 64 with a family history of osteoporosis, weight < 70 kg, prior nontraumatic fracture, and other risk factors
• Premenopausal women with therapy-associated early menopause
• Postmenopausal women of any age receiving aromatase inhibitors

Applying these general guidelines to female cancer patients, it would seem that two groups can be specifically identified as requiring BMD assessment: (1) patients < 45 years old who become postmenopausal because of their therapy, and (2) breast cancer patients who receive adjuvant aromatase inhibitors. Although for most of these women BMD determinations will be reassuring, such testing may identify patients at particular risk who may want to consider additional treatment to protect BMD and lessen their future risk of fracture. Interpreting DXA Scan Results
When used for screening, the crucial measure for the DXA scan is the T-score. T-score is an admittedly odd parameter, but easy to apply clinically. T-score reflects the difference between the patient's bone density and that of an average healthy young woman, in terms of standard deviations. The standard deviation is determined in a population of young women. If the patient has the same BMD as average healthy young women, the patient has a T-score of 0.0. If the patient has a T-score of -2, then her BMD is the same as that of a young woman whose BMD was 2 standard deviations lower than that of other average young women. T-scores gradually fall as women age, averaging -1.0 for women in their 60s, and -2.5 for women in their 80s.[25] When reviewing the DXA report, it is generally not necessary to calculate the T-score from the actual BMD (measured in g/cm3), as it is usually already calculated on the report. When T-scores vary at different sites (eg, discordant between the lumbar spine and hip), the least favorable (most negative) value should be used for treatment decisionmaking. If one of the sites is affected by severe osteoarthritis, data from this site should be discarded. T-scores are used for defining the state of normality of the BMD. The World Health Organization diagnostic criteria are shown in Table 3.[26] Guidelines recommend that all women engage in lifestyles and have a diet that is conducive to the maintenance of good bone health.[4,27] It is widely recommended that all women should maintain ample intake of vitamin D (400-800 U/d) and calcium (~1,200 mg/d) and remain physically active with regular weight-bearing exercise. Because the additional pharmacologic treatments used to prevent osteoporosis and fracture are only partially effective and involve cost and toxicity, the decision to use them involves balancing the expected benefit against cost and toxicity. Clearly, not all women need to be treated. Several guideline committees have addressed the question of who should be considered for treatment. A fairly representative set of guidelines (from the American Academy of Clinical Endocrinology) is shown below.[8]

• Women with postmenopausal osteoporosis-specifically, women with BMD T-scores of -2.5 and below and women with low-trauma fractures and low BMD
• Women with borderline-low BMD (eg, T-scores of -1.5 to -2.5) if risk factors are present
• Women in whom nonpharmacologic preventive measures are ineffective (bone loss continues or low-trauma fractures occur)

These recommendations are similar but not identical to those of ASCO, which recommended treatment for a more restricted set of women-that is, women with postmenopausal osteoporosis, or women with BMD T-scores of -2.5 and below. Although the ASCO guidelines only recommend starting therapy after patients actually became osteoporotic, they are in keeping with the results of the Fracture Intervention Trial, in which only women with a BMD < -2.5 had a significant reduction in clinical fracture risk with treatment (alendronate [Fosamax]).[28] Similarly, in the Hip Intervention Program Study, a significant reduction in the risk of fracture was observed with treatment (risedronate [Actonel]) only in women with confirmed osteoporosis.[29] Which Treatment for Osteoporosis Should Be Selected? The selection of additional therapy for osteoporosis may be different for breast cancer patients and women without a history of breast cancer. That said, treatment with bisphosphonates is one possible front-line therapy for women in the general population,[30] and these agents may also represent the most logical front-line treatment for many breast cancer patients. Their use does not raise the complex issues surrounding the use of agents that interact with the estrogen receptor, such as estrogens(Drug information on estrogens) (as used in hormone replacement therapy) or raloxifene(Drug information on raloxifene) (Evista). Bisphosphonates
Two bisphosphonates have been approved in the United States for the prevention and treatment of BMD loss: alendronate and risedronate. Both of these agents cause an increase in BMD in most postmenopausal patients.[ 3,29] Both agents are taken orally, on either a daily or once-aweek dosage schedule. Three years of treatment with alendronate has been shown to reduce the incidence of both vertebral and hip fractures by ~40%.[31] Similarly, 3 years of treatment with risedronate has been shown to reduce vertebral fracture risk by ~40% and to have a more modest beneficial effect on nonvertebral fractures.[ 32,33] No head-to-head comparison of these two treatment options has been conducted, and as studies have been carried out with somewhat different patient populations, there is no definitive difference in efficacy between them. Very few data specifically address the effectiveness of these agents in cancer patients, or whether they are effective when given together with agents used to treat breast cancer. There are no strong theoretical reasons to believe that they would be any less effective or have unfavorable interactions with agents to prevent or treat breast cancer. In one small randomized trial, it was demonstrated that risedronate helped prevent loss of BMD caused by the aromatase inhibitor letrozole.[ 34] There is also evidence that bone mass was improved in breast cancer patients with the use of the oral bisphosphonate clodronate (not yet available in the United States).[35] Conflicting but promising data suggest that clodronate might actually have adjuvant effects by decreasing recurrence and improving outcomes. The use of clodronate vs placebo in the adjuvant therapy of breast cancer is being tested in the National Surgical Adjuvant Breast and Bowel Project (NSABP) B-34 trial (http:// www.clinicaltrials.gov). Agents That Interact With the Estrogen Receptor
Although estrogens have been included in the list of options to prevent and/or treat osteoporosis for women in the general population, enthusiasm for their use is waning. This is due to strengthening evidence that their use is associated with an increased risk of cardiac disease and breast cancer.[36] They have long been considered contraindicated in women with a history of breast cancer because they probably increase the risk of recurrence (for women with a history of estrogenreceptor- positive breast cancer) and second primary breast cancers (for all women with a history of breast cancer).[37] Evidence from recent randomized trials is consistent with a major increase in the risk of recurrence among breast cancer survivors.[38] Tamoxifen is not indicated for the treatment or prevention of osteoporosis. In postmenopausal women, however, tamoxifen may help preserve BMD.[16] In the large BCPT, this agent reduced the risk of fracture in postmenopausal women by 20%, compared with placebo. Nevertheless, it should be noted that established osteoporosis treatments reduce the risk of fracture by a much greater amount (40%-50%). Fracture risk reduction was not a primary end point of the BCPT, and thus, although there is evidence of a mild protective effect of tamoxifen on bone, its benefit in a population of patients with established osteoporosis/osteopenia is unknown. In addition, a major concern about the use of tamoxifen for this purpose is that when used for more than 5 years, evidence shows that it may begin to exert estrogenic effects on breast cancer cells and no longer reduce or perhaps even increase the risk of recurrence and second primaries. There are arguments for and against the use of raloxifene to treat women with a history of breast cancer. In trials designed to test the effectiveness of raloxifene in reducing risk of breast cancer in postmenopausal women with osteoporosis, raloxifene-treated women had > 50% lower risk of developing breast cancer.[39] However, assessment of breast cancer risk was not a primary end point in this trial. Raloxifene has been shown to significantly decrease the risk of vertebral fractures,[23] but not nonvertebral or hip fractures.[40] The reduction in fracture risk was anticipated because raloxifene and tamoxifen both interact with the estrogen receptor and have many of the same properties. However, this similarity raises some important concerns, which are based on the limitations of tamoxifen. First, it is recognized that tamoxifen can paradoxically act as both an estrogen and an antiestrogen. There appears to be a paradoxic shift after long continuous treatment, whereby breast cancer cells are actually stimulated by tamoxifen. At least in animal models of breast cancer, this shift to an estrogenic-like stimulatory effect is also demonstrated by raloxifene after tamoxifen treatment.[41] This suggests that raloxifene would be a particularly poor choice after longterm treatment with tamoxifen, and that even the effectiveness and safety of long-term raloxifene might be called into question. Finally, the concurrent use of an aromatase inhibitor and raloxifene would seem to be contraindicated. Tamoxifen seemed to blunt the effectiveness of anastrozole in the ATAC trial,[10] presumably because of the estrogenic-like effects of tamoxifen; there are theoretical reasons to believe that raloxifene might have similar effects. Other Options
Other options for the treatment of osteoporosis and/or the maintenance of BMD are available, such as calcitonin and parathyroid hormone analogs. However, these are generally considered to be second- or third-line treatment options due to issues of expense, ease of administration, and/or toxicity. ASCO Guidelines
Agents specifically recommended in the ASCO guidelines were alendronate, risedronate, zoledronic acid(Drug information on zoledronic acid) (Zometa), or raloxifene. The guidelines made clear that these recommendations were based on effects in healthy women in the general population, not breast cancer patients. Some investigators have had theoretical safety concerns about raloxifene (as highlighted above). In addition, it was noted that the data supporting the use of zoledronic acid were based on very short-term and as yet incompletely reported trial results.[42] Follow-up After Screening or Treatment for Osteoporosis With normal aging, both women and men typically lose 1% to 2% of bone mass per year. A loss of ~10% to 15% equals one T-score unit change. Many guideline committees suggest repeat evaluations every 1 to 2 years or as clinically indicated. For individuals with excellent and stable BMD, and no change in treatment or endocrinologic status, it would seem reasonable to consider a longer interval. Measuring BMD at intervals shorter than 1 year is clearly not clinically appropriate because of the limitations of BMD testing. There are typically ~1% standard errors in these measurements (this error parameter should be on the BMD report) and a 95% confidence interval is defined as 2.8 standard errors. Thus, changes in BMD of less than ~3% on many machines are often due to measurement error and should not be used to influence clinical decisions. Given the usually modest drift in BMD of 1% to 2% per year, it is clear that at least 1 year is needed to be sure that a change is occurring. Large clinical trials can report changes in composite BMD estimates that are smaller and at shorter intervals, but they are reporting averaged data from many individuals. For the greatest confidence in measuring changes, BMD measurements should be carried out on the same machine. Comparing BMD measurements taken at different times, different bone sites, or with different measurement methods cannot be used to accurately determine if changes in BMD are occurring. If the patient shows a fall in BMD or has a non-cancer-related fracture despite treatment, a change in therapy should be considered. Such patients might benefit from a referral to an endocrinologist with a special interest in osteoporosis, for a more indepth evaluation of other causes of the disease and a broader review of treatment options. ASCO Guidelines
ASCO guidelines recommend an annual clinical reassessment of osteoporosis for all women, even those at low risk. They also recommend that annual BMD measurements be carried out for women considered to be at high risk by history and physical exam. For the reasons stated above, these recommendations may mandate more BMD testing than would seem necessary for some patients. Conclusions Today many cancer patients and most breast cancer patients become long-term survivors. Thus, women with cancer, like women in the general population, need screening for risk of potential serious future health problems. One of the most common of these potential problems is osteoporosis, which if followed by fracture can cause severe disability and even a significant risk of mortality.[43] There are reasonably effective osteoporosis screening and treatment strategies that should be applied to all postmenopausal women and all female cancer patients. Some types of therapy, such as those that cause early menopause or that lower postmenopausal estrogen levels, might be considered significant risk factors for osteoporosis. For breast cancer patients identified as being at heightened risk of osteoporosis and fracture, bisphosphonates appear to be a logical choice for treatment. Agents that interact with the estrogen receptor pose certain theoretical risks. In addition, although tamoxifen has a modest positive effect on BMD and fracture risk, this benefit should not be interpreted as sufficient for the majority of postmenopausal breast cancer patients with low bone mass. In general, much more research is needed in this area as a number of questions remain unanswered. Many recommendations are based on shortterm outcomes; hence, more information about lifelong risks, benefits of surveillance, and treatment options would be of value.