BoneKEy-Osteovision | Meeting Reports
Meeting report from the national osteoporosis foundation 5th international symposium on clinical advances in osteoporosis
DOI:10.1138/2002038
Osteoporosis overview and global implications Dennis M. BlackThis session presented a general overview of the pathogenesis and problem of osteoporosis around the world. This review will focus only on a few of those presentations while briefly summarizing the remainder.
Summaries of first presentationsThomas Einhorn presented a succinct overview of the problem of osteoporosis, a summary of its pathophysiology and discussed determinants of bone strength.
Dennis Black, in an overview of bone mineral density, summarized the accumulating data that bone mineral density predicts fractures and that BMD at the hip is most useful for predicting hip fractures and general fracture risk. He summarized the recent work on peripheral densitometry, showing that most peripheral devices are at least somewhat predictive of fracture risk. However, use of T-score-based diagnostic and treatment thresholds from peripheral devices leads to huge systemic differences in the proportion of women (and men) identified as having osteoporosis and to inconsistency. In addition, the fracture risk among those with BMD below specific T-score thresholds varies dramatically from device to device. Therefore, he recommended that central DXA, particularly at the hip, be used for diagnostic and treatment decisions and cautioned against the use of T-scores from peripheral devices for diagnosis.
Xu Ling summarized recent data from Beijing studies confirming that hip fracture risk is lower in mainland China and that paradoxically, BMD is also lower. However, some of the differences in BMD disappear when adjusted for differences in body size. These data suggest that different absolute levels of BMD should be used for diagnosis of osteoporosis in China and other parts of Asia, although it is not yet clear how to develop these thresholds. It may be that use of T-scores derived from Asian reference data may represent an interim solution until more data are available relating BMD to fracture in this region.
Cyrus Cooper summarized the international epidemiology of osteoporosis and fractures. He noted that rates of both hip and vertebral fractures are similar in Europe and North America. However, he noted that while hip fracture rates were generally lower in Asia, vertebral fracture rates in Asia were similar to North America and Europe. Geographic heterogeneity is only partly explained by ethnicity. In attempting to predict the magnitude of the future international problem of osteoporosis, it is important to take into account the differing secular trends. For example, hip fracture incidence is estimated to have increased 3.3% per year in Hong Kong between 1966 and 1989 while only about 1% per year in the U.S. He said that our current projections do not take into account this varying “tempo” in the increase in incidence of fracture risk and that we urgently need to better understand these trends in order to project future incidence and develop better population strategies to deal with the problem of osteoporosis in the future.
Why do fracture rates vary? Role of genetics and environmentStuart Ralston gave an overview of what is known about the genetics of osteoporosis. He started by considering peak bone mass and concluded that majority of peak bone mass is not environmentally determined and is probably under genetic control. More specifically, in studies that included such environmental influences as exercise, height, weight and medication use, almost 80% of the variance remained unexplained by these environmental factors. Furthermore, studies of twins suggest that spine and hip peak bone mass is 70-80% heritable. As with peak bone mass, the majority of post-menopausal bone loss cannot be explained by environmental factors and is therefore likely to be genetic. There is some direct evidence for genetic control of axial bone loss. The data for bone markers suggest that, in general, they are less strongly heritable, with the exception of BSAP levels which seem strongly genetically determined. While there are few data, there is some suggestion that bone geometry and quality are highly heritable. In fact, a recent study by Arden and colleagues suggested that specific quantitative trait loci (QTL) exist for femoral geometry. On the other hand, some aspects of bone strength such as cortical thickness and cross-sectional area are strongly affected by exercise, suggesting a strong non-genetic element.
However, despite the evidence for strong heritability for bone mass, bone loss, markers and geometry, the direct heritability for fracture is less than expected. The heritability of fractures is only on the order of 30%.
Dr. Ralston described the role of the various types of genetic studies in osteoporosis. It seems likely that osteoporosis is a polygenic (not monogenic) disease and therefore association studies are less likely to lead to important results. Some light has been shed on osteoporosis genetics from family studies and from linkage studies in sib pairs. However, most productive have been studies in mice and an increasing appreciation of the overlap in the mouse and human genomes.
There are several specific genes that have been investigated for their relationship to osteoporosis. A few years ago, there was a flurry of interest in the vitamin D receptor gene (VDR). However, after a large number of studies, it is becoming clear that this gene plays a relatively small role. Interestingly, some recent work by Krall () and others has suggested the VDR gene may be involved in a gene-environment interaction in which an individual's response to calcium supplementation may be partly a function of their VDR genotype. More recently, there has been a developing interest in the Type I collagen (COL1A1) gene. This gene is not strongly related to BMD but may be related to bone quality and/or fracture risk. Studies have shown that COL1A1 is correlated to bone strength as assessed by mechanical testing. It may also be correlated to osteoblastic protein production. It is thought that COL1A1 may affect bone mineralization. One important role for genetics in osteoporosis may be to identify potentially new therapeutic targets for drug development. The fact that some high risk genotypes (e.g. COL1A1 ss) are quite common is very promising. Dr. Ralston summarized his discussion with the observation that we are only beginning to understand the genetics of osteoporosis: most genes remain to be discovered.
Should guidelines for BMD measurements be universal?In perhaps the most thoughtful presentation of the session, Joseph Melton discussed whether BMD guidelines should be universal. He noted the huge increase in the number of elderly in the world expected over the next 50 years and the consequent increase in fractures. And these projected increases in the number of fractures over the next 50 years may be underestimating the problem, since they assume that age-specific fracture incidence does not increase, while data from places such as Hong Kong suggest an increase in age-specific rates will accompany economic development.
However, while the problem of osteoporosis, particularly hip fractures, will greatly increase over the next 50 years, the incidence of other diseases of aging will also be accelerating in parallel. For example, the incidence of dementia and hip fracture with age are similar and thus will increase in parallel as the population of the world ages. Cancer will show similar increases. In addition, the worldwide burden of infectious disease and violence are dramatically increasing. As the overall costs of health care increase, there will be a competition for increasingly limited health care resources. Osteoporosis is not even among the 10 diseases with the highest societal burden and therefore we can expect it to be increasingly ignored. In this climate, particularly in the developing world, there is a need for a public health approach to osteoporosis and a need to develop strategies for prevention and treatment which are inexpensive, safe and can applied across a broad range of the population.
Anti-fractures efficacy of therapeutic agnets Felicia Cosman The bisphosphonatesSusan Greenspan reviewed clinical trials using the bisphosphonates etidronate, alendronate and risedronate. She pointed out that the drugs cannot be compared directly with each other because there are many differences among the clinical trials. The differences include severity of the osteoporosis at baseline, differences in fracture rates in placebo groups, definitions of outcomes with regard to both vertebral and non-vertebral fractures, and statistical analyses. For example, vertebral fractures were sometimes new, sometimes new and worsening, sometimes defined as 20% loss of vertebral height, sometimes 15%. Data concerning primary outcomes were reviewed. Drugs were evaluated in terms of what is the target population, how fast the medication works, how long the duration of data on the drug, and whether the drug worked on spine as well as hip. They include studies where applicable in men, postmenopausal women and steroid induced osteoporosis. A summary of the data is shown below:
Non-skeletal effects of estrogenElizabeth Barrett Connor reviewed the non-skeletal effects of estrogen, including those on cardiovascular disease, breast, cognitive function and quality of life. She concentrated on cardiovascular outcome studies. These included HERS (), ERA, PHASE (), WEST (), EAGAR, EPAT, PHOREA, and EPAT. In the HERS study, there was a significant increase in risk of cardiovascular disease outcomes in the first year, of about 50%. Overall, the results after 4 years indicated no overall benefit against cardiovascular disease. A recent publication in Circulation () evaluated subgroups post hoc to see if any could be identified who benefited or were harmed by HRT in HERS. There were no assumptions made a priori to support the subgroup classification. Of 172 tests for interaction, there were 9 positive findings (consistent with chance).
While some have commented that the lack of effect of HRT on heart disease outcomes in HERS may have been related to the use of progestin plus estrogen, rather than estrogen alone, HERS is not the only clinical trial where no CVD benefit was found (). In ERA, there was no reduction in progression of angiographic disease in those treated with estrogen alone or estrogen plus progestin for a 3 year period. Furthermore, some have argued that perhaps different estrogens might produce different effects than oral Premarin. In fact, the PHASE () study utilized transdermal estradiol and still did not produce beneficial cardiovascular outcomes. The WEST () study for secondary stroke prevention utilized oral 17-beta estradiol. Some have argued that these negative results have been only in secondary prevention studies; however, the Women's Health Initiative (WHI) () has also indicated that an early increase in cardiovascular, cerebrovascular disease and venous thromboembolism has been seen in estrogen treated women in this primary prevention trial. 30% of women in WHI are on unopposed estrogen. The American Heart Association's recommendation to renounce its support of estrogen for secondary prevention against heart disease remains the correct decision.
She commented about the estrogen-breast cancer relationship being based solely on observational data, but that the Collaborative Study (a metaanalysis) indicated an increase of about 2.3% per year beyond 5 years of use. Information from WHI should help answer the breast cancer question.
Recent data from HERS also suggest that estrogen does not improve quality of life unless women taking it have hot flashes when they begin. Estrogen does not improve incontinence. There are no clinical trial data that it reduces dementia risk or severity. It does increase risk of venous thromboembolism and gallbladder disease.
Estrogen and bone lossThe biology of estrogen deficiency and bone loss was reviewed by Robert Lindsay. Clinical trials with bone density outcomes have consistently shown maintenance or improvement in bone density in all populations. These include older studies, such as those by Lindsay and the larger PEPI study (). Low dose estrogen was compared with standard dose estrogen in the recently presented HOPE study and bone density changes were the same with all doses of estrogen.
Almost all of the fracture data with estrogen are observational and they suggest that estrogens can reduce fracture risk by about 50%. Clinical trials with fracture outcomes were reviewed, including the Finnish study () and Danish study () in early postmenopausal women. The first study indicated that estrogen could reduce clinical fractures compared with placebo (but produced an effect similar to that of Vitamin D) and the latter suggested that estrogen could reduce wrist fracture occurrence. The Torgerson metaanalysis of clinical fracture occurrence in RCTs of estrogen () was also reviewed. All randomized clinical trials with any fracture data recorded were included; some were non-published. The metaanalysis indicated that estrogens could reduce fractures in early postmenopausal women. In older women, the analysis was dominated by the HERS trial and therefore the results in older women were not positive. In a group of women with osteoporosis, however, Lufkin () showed a reduction in number of new vertebral fractures. Furthermore, in a smaller group of women with heart disease similar to those in the HERS trial, fracture rates were lower in the estrogen groups compared with placebo.
SERMS, tibolone and calcitoninTwo SERMS currently used in healthy postmenopausal women were discussed by Felicia Cosman. Tamoxifen is approved for prevention of breast cancer but also has skeletal effects, and raloxifene is approved for prevention and treatment of osteoporosis. Tamoxifen improves bone density modestly in postmenopausal but not premenopausal women, where its estrogen antagonist effects appear to predominate. The only fracture data are from the Breast Cancer Prevention Trial, a study of 13,328 Women (both premenopausal and postmenopausal) who were at elevated risk of breast cancer (). The total number of fractures was the same in tamoxifen and placebo groups. The number of hip, clinical spine and wrist fractures, however, was about 19% lower in tamoxifen treated patients, a difference which did not meet statistical significance. This modest difference may be related to several factors: that the population was low risk, that the fracture recording was not comprehensive, or that there were no spine xrays or height measurements. Tamoxifen also reduces risk of breast cancer by about 50%, does not have a significant overall effect on cardiovascular or cerebrovascular disease, but does increase risk of venous thromboembolic disease and endometrial cancer.
Raloxifene reduces bone turnover and improves bone density in early postmenopausal women in prevention studies and in women from the MORE trial in 7705 women with osteoporosis (). It significantly reduced risk of morphometric vertebral fracture by 33-55% (in the subgroup of women with prevalent vertebral deformity at study entry, and in the subgroup without prevalent vertebral deformity) at 3 years with sustained effects at 4 years. Raloxifene also reduces overall risk of breast cancer by 65%, risk of estrogen receptor positive breast cancer by 85% and has no effect on estrogen receptor negative breast cancer. Breast cancer risk in placebo treated women was highest in women with highest baseline estradiol levels and raloxifene had its biggest effect in this group. Recent cardiovascular disease outcome data from MORE were also presented. There was no early increase in risk of cardiovascular disease and no overall effect in the entire population at 4 years. In women with known cardiovascular disease or those at high risk for heart disease (n=1035), there was a 40% significant reduction in risk of combined fatal and nonfatal cardiac and cerebrovascular outcomes (). Raloxifene does not increase uterine disease, has a neutral effect on cognitive function at 3 years, does not increase pelvic organ prolapse (and might even reduce need for pelvic floor surgery), but does increase risk of venous thromboembolism (similar to risk for estrogen and tamoxifen).
Tibolone is a drug marketed in other countries but not yet in the US. It reduces hot flashes and sweats similarly to estrogen and might improve irritability and libido in postmenopausal women. While it improves bone density and reduces bone turnover, effects on fracture are unknown. Tibolone does not increase breast density or produce breast tenderness, but effects on breast cancer are unknown. It may produce some dose-dependent uterine bleeding, although less than seen with estrogen. Effects on lipids are variable and influence on vascular disease is unknown.
Nasal calcitonin has been studied most completely in the PROOF study (), which enrolled 1175 women with osteoporosis, 80% of whom had vertebral compressive deformities at baseline. The very high drop-out rate in this study, about 60%, limits the conclusions that can be drawn. Changes in bone density were modest and not significantly different from placebo at years 3-5. Changes in bone turnover were only different from baseline at year 1. Incident vertebral deformity was reduced by 33% in women receiving the 200 IU dose, but not at either of the other doses (100 IU, 400IU). There were no significant differences in fracture occurrence at other sites.
Guidelines for osteoporosis Clifford J. Rosen BackgroundIt has been five years since the National Osteoporosis Foundation released guidelines for the diagnosis and treatment of osteoporosis. Prior to that time, the WHO had provided clinicians with a useful classification of bone mineral density measurements based on the strong inverse relationship between units of standard deviation below young normal, and percent of women with osteoporotic fractures. In 1997 the NOF sought to clarify both the diagnostic and therapeutic aspects of various BMD cut points. At the time of their introduction these guidelines, which included cost effectiveness analyses (i.e. quality-adjusted life years saved) and risk factor evaluation, were surrounded by controversy, especially among specialists in metabolic bone disease. However, at least in the U.S., the NOF guidelines, published first in Osteoporosis International (), and then as a small handbook, soon became a useful tool for primary care physicians. The NOF recommended that women with T-scores of the hip lower than -2.0, or -1.5 with at least 1 clinical risk factor (smoking, previous fracture, weight, maternal history of hip fracture), or women over the age of 75 with multiple clinical risk factors, should be treated aggressively to prevent fractures. At the time, however, only calcitonin, alendronate and HRT were approved treatments for osteoporosis. Moreover, no consideration was given to men with osteoporotic fractures, men or women with glucocorticoid induced osteoporosis to peripheral bone mass measurements or the global issue of prevention.
SymposiumAt the 2002 Hawaiin NOF symposium, Robert Lindsay revisited the issue of guidelines. Since 1997 new issues have arisen, new trial evidence has become available, new treatments are now accessible, and multiple investigators have produced risk profiles. Thus, a major theme of this session was that the thresholds for diagnosing osteoporosis by BMD differ from those for treating individuals based on BMD. Dr. Lindsay first addressed the issues of peripheral densitometry as predictors of fracture risk. The NORA study () has unequivocally shown that peripheral measurements predict risk of osteoporotic fractures as well as central determinations, although hip BMD remains the best site for predicting hip fracture. Surprisingly, the NORA study revealed that the actual number of fractures among participating women was slightly higher than expected and some fractures were noted in women with T-scores between 0 and -2.0. Hence, concern has been raised that a -2.0 cut point for peripheral BMD measurements may not catch the majority of women at high risk. It has also become clear from NORA and other studies that BMD alone may not be the best predictor of risk in older populations. Evidence for that emerged last year from Lindsay et al. (), who reported in a large cohort of women with osteoporosis that there was a 20% risk of a new vertebral fracture (accompanied by significant morbidity and mortality) within a year of a incident vertebral fracture. Data from these studies and others suggest that absolute fracture risk, rather than relative fracture risk, is a more appropriate means for expressing overall likelihood of subsequent fracture. Hence, BMD plus other clinical risk factors must be considered, especially in older women.
Cost effectiveness for a specific osteoporosis treatment has become a more complex issue than it was five years ago, in part because of the greater number of therapeutic options (and combinations) as well as the absence of head to head studies between agents of the same class (e.g. bisphosphonates). Using large multicenter trials with individual agents vs placebo and systematic reviews, there is sufficient evidence to support a specific “number needed to treat (NNT)” for both prevention and treatment. However, based on the above considerations, no new recommendations were issued at this meeting in respect to therapy.
Dr. Lindsay also addressed the contentious issue of prevention. Since the original NOF guidelines, several studies have been published that suggest preventive therapy with bisphosphonates in women with T-scores <-2.5 may reduce the risk of fracture. In FIT-2 () women with T-scores < -2.5 in the hip who were treated with alendronate had a relative risk reduction of approximately 36% in non spine fractures (i.e. the NNT was 15 to prevent one non spine fracture). Similarly in the FOSIT study (), women with T-scores <-2.0 who received alendronate had significant risk reduction for non spine fractures (RR = 0.78; NNT 24 to prevent one non spine fracture). For vertebral fractures, in FIT-2, women with T scores <-2.5 treated for three years with alendronate had a 50% risk reduction for new spine fractures; those between -2.0 and -2.5 had a 46% risk reduction; and those with a hip BMD greater than -2.0 who underwent treatment, had no effective risk reduction. The conclusion Dr. Lindsay reached was that prevention in women with T scores between -2.0 and -2.5 could be accomplished with the alendronate.
Dr. Lindsay remains confident of the following recommendations: 1) you should perform BMD on all postmenopausal women at least once to assess risk; 2) you should treat women with T- scores <-2.5 at any or all sites; 3) you should treat postmenopausal women with T-scores higher than -2.5 if there are clinical risk factors.
In the second part of the symposium, Eric Orwoll presented displayed the initial efforts to come to a consensus on guidelines for the diagnosis and treatment of osteoporosis in men. Over the past two years, Dr. Orwoll, under the sponsorship of the American Society for Bone and Mineral Research, has been working on a modified evidence based approach to male osteoporosis. This is a three step process: 1) a comprehensive literature search; 2) polling international experts with a series of questions about diagnosis and treatment; 3) convening a consensus panel to prioritize responses and provide a guideline statement. One of the most intriguing aspects of the presentation was the preliminary results from step 2. Twenty four world experts on osteoporosis in males responded to 160 questions and scored their response from 1-10 (1-3 not recommended; 7-9, strongly recommended; in the middle- 5-6- no consensus). As for BMD measurements, screening all men for osteoporosis scored very low, while most experts agreed (approximate score of 8) that men with previous fractures, men on glucocorticoids, men with hypogonadism, men post transplant, and those with malabsorption should have determinations of BMD. Most experts strongly recommended that the T-score cut-off point for treatment be considered -2.5 or -2.0 in men with prevalent fractures. Bisphosphonates were highly recommended whereas calcitonin, SERMs and estrogen were given very low scores. Central, but not peripheral measurements were endorsed. The final consensus recommendations for male osteoporosis await further analysis of the data, as well as completion of the questionnaire by other investigators.
The last part of the session dealt with an issue which did not arise in the first publication of guidelines; i.e. ethnic reference sets and the appropriateness of universal T-scores. Edith Lau was very clear that for Asian populations, risk based on BMD is not clear. In part, this is related to the disparity between hip fractures in women with low BMD in Asia versus those in North America (i.e. Asians, at least some subsets including Chinese, have lower hip fracture rates for the same BMD as Caucasians). On the other hand, lifetime vertebral fracture risk is similar if not slightly higher in Asians, than Caucasians (30% vs 25%). Complicating this issue is the fact that volumetric BMD (BMAD)which corrects for size of the bone, does not differ between Asian postmenopausal women and Caucasians. Interestingly, Dr. Lau also noted that Asian women are less likely to take HRT (i.e. only around 3% are currently on hormone replacement). But the response to alendronate does not differ by race. Finally, Dr. Lau noted that preliminary studies suggest that Tai Chi in Asian women can prevent postmenopausal bone loss.
Carlos Mautalen discussed differences in bone density and fracture risk among Hispanic women. He emphasized the very strong regional differences within countries such as Argentina for hip fractures and Mexico for BMD. He made a strong case that application of U.S. NOF guidelines to Latin America is not practical at this stage, and that further research is needed.
ConclusionsAlthough great strides have been made over the five years since the publication of the original NOF guidelines for treating osteoporosis, the field remains divided about several issues. In particular, the use of T-scores, the application of absolute fracture risk, the threshold for treatment, the use of cut-offs for aggressive intervention with bisphosphonates, and the use of markers or BMD to follow patients continue to be controversial. Those issues were highlighted at this meeting, and are largely responsible for the absence of a new NOF- guideline for treating osteoporosis. On the other hand, standards for diagnosing and treating male osteoporosis are likely to be published in the next year, and more evidence-based approaches to treatment using systematic reviews are clearly warranted.
Monitoring response to osteoporosis treatments: Viewpoints Richard Eastell Treatment should be monitored with BMDNelson Watts described the need to measure the response to treatment; since we cannot evaluate fracture risk within individuals directly, it is appropriate to measure either bone mineral density (BMD) or bone turnover markers (BTM). He felt that BMD gave an opportunity for the ‘personalisation of an abstract generality’ thus attempting to win the debate by the use of polysyllabic words. He went on to describe the standard approach to identify patients who respond by characterising the variability of bone density measurements over the appropriate period of time and calculating the least significant change. He pointed out that this approach identified responders to HRT and bisphosphonate therapy (although few responders to raloxifene and calcitonin therapy), but that it was important that careful attention be paid to the details of obtaining consistent results.
Watts acknowledged that there are limitations to monitoring with BMD. The change in BMD accounts for less than 50% of the reduction in fracture risk with antiresorptive therapy, and so it is not an ideal surrogate. The initial change in BMD may not predict subsequent change in BMD as a result of regression towards the mean. He quoted work suggesting that regression towards the mean is a group phenomenon () and presented his own analysis () suggesting that regression towards the mean has little impact on the classification of individuals in practice.
He finished his presentation by quoting some figures from an audit of BMD practice in Western Australia that showed 15% of patients with osteoporosis followed longitudinally had bone loss beyond the least significant change, and thus need to be identified and treated appropriately - the treatment changed or the secondary osteoporosis identified.
Don't monitor treatment with BMDSteven Cummings presented four arguments against the use of BMD and concluded by saying that the main value was the psychological reassurance given to the patient and physician. First, the follow-up BMD result is taken too late to be of any use. Patients usually stop taking treatment in the first year - 76% of patients stopping alendronate in the Fracture Intervention Trial (FIT) did so in the first year. Second, bone loss does not imply no response to therapy - patients may have lost much more had they not taken the treatment. There is only a weak association between pill count in a clinical trial and BMD response. Third, the change in BMD during the first year is a poor predictor of subsequent BMD change. For example, for women losing more than 4% a year or more on alendronate, 92% will gain bone during the following year. This observation is true for other agents, e.g. raloxifene, and for the hip and the spine (). Fourth, patients with bone loss while taking a bisphosphonate may still have a reduction in their fracture risk. In the FIT study, the 18% of women with BMD loss taking alendronate had a fracture rate half of that of the 18% women on placebo with the greatest bone loss.
Treatment should be monitored with markers of bone turnoverDr. Siris introduced Richard Eastell as the ‘Brit with the Mitt, the Sheffield Stomper’! This changed the temperature of the water and turned a viewpoint into a debate. Eastell pointed out that in clinical practice it was common to observe secondary osteoporosis and to come across patients who do not adhere to the therapeutic regime, nor do the noncompliant always tell the physician. He felt that bone markers had the advantage over BMD in that the change in bone resorption markers is maximal at 3 months and is very large. Further, there is recent information suggesting that BTM are more closely associated with fracture risk reduction than BMD. He used the observation that fracture benefit with risedronate has been shown as early as six months, long before the maximal BMD benefit, and there is an association between change in BTM (osteocalcin) and vertebral fracture risk reduction with raloxifene () but not with change in BMD. He used some unpublished results showing that changes in urinary N-telopeptide (NTX) and C-telopeptide (CTX) explain about 66% of the vertebral fracture risk reduction with risedronate, and that a 50% decrease in these markers was associated with least risk of fracture.
The weakness in the case for BTM for monitoring was in their variability and Eastell recommended care in taking bone resorption markers at a fixed time of day and to obtain two results before starting therapy (and taking the average as the baseline value). Using this approach and a least significant change value for urinary NTX of 54% we find that with standard antiresorptive treatment, about 60% of patients will be considered responders and this can be identified within the first six months of treatment. The measurement of these markers has improved with the introduction of automated immunoassay analysers and point of care devices.
Markers don't help in monitoring treatment of osteoporosisMarc Hochberg was introduced as the Man with the Mantra that Markers Don't Help, ‘Iron Marc Hochberg’! He begin a little mischievously by misquoting an advocate of BTM, Dr. Delmas () as saying ‘Their clinical use in the management of the individual patient has not been clearly defined and is a matter of a debate.’ However, that was the introduction to a position paper, not its conclusion! He felt that a significant weakness to the use of BTM in clinical practice was their variability and described the sources of variability using the work of his opponent () and stating that this variability made the markers impractical to use - when challenged during the discussion period he felt that patients would not attend for more than one baseline sample and that patients were not prepared to attend the clinic in the fasting state in order to reduce variability. He referred to an abstract at the meeting (Abstract 14) that presented a meta-analysis of the usefulness of BTM. This concluded that high bone markers were not sufficiently predictive of rapid bone loss for use in the individual. He also felt that the identification of a non-responder had little practical benefit as most drugs were approved for use at a single dose and so dosage adjustment was not possible. Finally, he felt that adherence was not improved by feeding back to the patient the results from the marker measurements. He based the latter conclusion on an unpublished study by Silverman of 240 patients with osteoporosis, mean age 67 years, who were prescribed alendronate and had usual care, nurse education, feedback of NTX results from baseline and 3 month measurement, and both nurse education and feedback. He found that the feedback of the NTX measurement had no effect on persistence with treatment, although there was a positive association of compliance with the change in BMD.
DiscussionQuestion: How common is secondary osteoporosis? SC claimed this was uncommon in primary care and may not attenuate the response to treatment - he gave the example of the benefit of alendronate to men even if they are hypogonadal. RE used the example of the absence of response to bisphosphonate in the setting of vitamin D deficiency ().
Question: the least significant change approach assumes a p-value of 0.05, but physicians would be happy with a p-value of 0.10. What difference would this make to the least significant change? RE said that the biggest effect on the size of the least significant change was to obtain two (or more) samples and take the average before estimating change. The effect of reducing p-value from 0.05 to 0.10 was 16% (e.g. for BMD a least significant change would go from 2.77% to 2.33%).
Comment: feeding has a large effect on bone resorption markers (). Thus, the reason why some studies but not others are able to show a strong correlation of change in markers to change in BMD may be a result of the care with which the marker samples were obtained.
Vote A vote was taken at the end of the session. Each speaker received approximately the same number of votes - there was no count, so we will never know who really won. We suspect this is a debate that will run and run!
Non-pharmacologic management today and pharmacological treatment tomorrow Gregory R. MundyThe final session of the 5th International Symposium was devoted to life-style factors important in the management of osteoporosis and new treatments near and far. The first presentation was given by Christine Snow, who discussed exercise and hip pads in the management of falls. Falls in the elderly are clearly important for osteoporosis, since 90% of hip fractures are associated with falls and falls in old people are tremendously common. Dr. Snow suggested that there is one fall per second in the elderly population in the US, and although most are not associated with hip fractures, hip fracture occurs almost always in association with a fall. Falls to the side with impact at the hip are the most dangerous for hip fracture. She reviewed the well-known risk factors that are associated with falls, including cognitive and visual problems, as well as extrinsic factors such as icy weather and medications that predispose to falls.
The most important part of this presentation was devoted to the importance of exercise and strength training in the prevention of falls. Dr. Snow reviewed the literature and described new data of her own. Gait, balance, and lower extremity strength are determinants of risk for falling, and strength training in the elderly has potential as an intervention to prevent falls, particularly exercise to improve strength and lean muscle mass. Improvement in balance and psychological well-being are important secondary benefits. She discussed the reported exercise intervention studies showing that strength training reduces risk of falls and improves balance, and concluded that exercise can therefore be assumed to reduce the risk of hip fracture. Detraining leads to reversal of these beneficial effects.
The results are clear and the benefits obvious. What is not entirely clear is how exercise as an intervention is best introduced into an osteoporotic clinical population. This was not addressed by Dr. Snow. Who is best to advise, and how is it best done and monitored-physical therapists, nurses, physicians, personal trainers?
Dr. Snow concluded her discussion with a brief discussion of hip pads and their potential benefits. She noted studies, mostly quite small, suggesting reduction in peak impact forces and reduction of hip fracture rates by 50-60% in at risk people. She noted compliance may be a problem with hip pads, but did not indicate how practical this intervention is, and when it should be considered in a patient in an osteoporosis clinic. Is it still experimental, or is it ready for prime time?
There were several interesting poster presentations that pertained to Dr. Snow's presentation. Gustavsson and Nordstrom (Abstract # 166) showed that bone mineral density was greater in adolescent male ice hockey players, but was not maintained unless they continued to play active ice hockey, again emphasizing the importance of continued load-bearing on maintenance of bone mass. Abstract #108 by Waltman et al. examined in a pilot study the feasibility of strength training for improvement in bone mineral density after breast cancer surgery, giving some indication of the size study needed to gather useful information of this sort.
The second presentation was Connie Weaver on dietary constituents and vitamins. Dr. Weaver gave a strong endorsement for the value of dairy products for bone health, suggesting that bioavailable calcium and other critical constituents in milk make it the most desirable functional food for osteoporosis. She noted the particular importance of dietary calcium in adolescent females and in achieving appropriate peak bone mass. She was questioned about evidence of its value in longer term studies compared with short term studies, where the data are less convincing. She also briefly reviewed dietary magnesium, important in the mineralization of bone, and potassium, which suppresses calcium excretion in the urine. Sodium (or dietary salt) promotes urinary calcium excretion and urinary sodium correlates negatively with hip bone mineral density. She noted a number of studies suggesting vitamin D intake is inadequate in the elderly, and attention needs to be paid to this. Protein intakes which are too high or too low are bad for the skeleton. She suggests a diet rich in dairy products, fruits, and vegetables is ideal. A relevant abstract was #128 by Segal et al. from Israel, who examined bone mineral density in lactase-deficient individuals, and found that lactase deficiency is accompanied by calcium and vitamin D deficiency and is associated with increased bone resorption and low bone mineral density.
Sundeep Khosla discussed so-called functional food constituents, particularly phytoestrogens. He noted that Asians have a high intake of soy protein and relatively high blood levels of phytogestrogens compared with Caucasian populations. A number of animal studies suggest beneficial effects on bone. Khosla discussed whether these effects were due to binding of these plant products to estrogen receptors or occurred independent of the estrogen receptor. Very few human data have been published, but the rat data clearly show a beneficial effect of some flavanoids on bone. He also reviewed studies with the synthetic flavanoid Ipriflavone, describing a recent study showing no effect in humans at doses that cause lymphopenia.
This area still remains uncertain. However, it is striking that: (a) biologic effects of phytoestrogens are obvious, (b) there are no definitive studies in humans, (c) the appropriate dose and mechanism of action are unclear, and (d) effects may be even greater than those of traditional functional foods or additives such as milk, calcium, and vitamin D.
The next presentation was given by Robert Neer on anabolic agents. He reviewed briefly the current status of all anabolic therapies but focused on PTH, with which he has been associated for several decades. He reviewed the critical human studies that show PTH has an anabolic effect on the skeleton, and has a better effect on fracture rates and bone mineral density in large studies than any of the anti-resorptive agents. The data are now well known and are indeed impressive. Dr. Neer concluded his presentation by a personal account of how he thinks that PTH might be used in clinical practice should it be approved by the FDA. He felt that it should be reserved for those high risk patients who have established osteoporosis, prior fragility fractures, low bone mineral density and who suffer from frequent falls. The cost of this drug still remain unclear, but it will be expensive. The possibility of its use in combination with other agents, particularly anti-resorptives, remains uncertain. Abstract #158 and #159 were relevant to this presentation. From the same group in China (Zhang et al.), it was found that PTH causes an increase in IGF-1 expression and production by osteoblasts in vitro, which confirms earlier observations, and recombinant human PTH increases trabecular bone volume in rats, and has greater relative affects on trabecular bone than on cortical bone.
The final presentation was given by Gregory Mundy. He discussed new treatments on the horizon and beyond. He indicated the ideal agent for osteoporosis when it becomes available will be the drug that not only prevents bone resorption and stimulates bone formation so that it can be used for prevention as well as treatment, but is also oral, safe, inexpensive, and convenient, with desirable pharmacokinetics. He reviewed the importance over the next two decades of rational drug discovery using molecular targets that are important in bone resorption and bone formation for drug discovery and made several suggestions as to what these targets might be, including critical transcription factors such as Cbfa-1, molecules in the receptor-signal transduction cascade of growth factors that regulate bone resorption and bone formation, or even growth factor ligands themselves. He described using the promoter of bone morphogenic protein 2 (BMP2) as a target for drug discovery to show that statins have stimulatory effects on osteoblasts and then discussed other studies arising from these observations, including the importance of the mevalonate pathway on bone resorption and bone formation and inhibitors of the ubiquitin-proteasome pathway to regulate critical molecules in the BMP2 signal transduction cascade leading to increased bone formation. The future is rosy for the development of new agents, but there are many hurdles involved in drug discovery in the field of osteoporosis, including prolonged development time, the great costs in bringing drugs to market in this particular area, and inadequate parameters as surrogates for assessing fracture rates. He also mentioned how current patent law may be a major hurdle for some drugs because of the long development times that are required.
This work is licensed under a
Creative Commons Attribution-Noncommercial-No Derivative Works 3.0 United States License.