Use of Bone Turnover Markers in the Real World: Are We There Yet?

Early identification of the patient at risk for osteoporosis is pivotal for the effective prevention and management of a condition that, if unrecognized, almost invariably deteriorates, leading to considerable complications such as fractures, chronic pain, immobility, social isolation, and death. Over the years, we have learned that, in addition to the well-known clinical risk factors of osteoporosis, accelerated bone turnover is associated with a higher risk of fracture, independent of age, sex, body weight, disability, and bone mass. Hence, the rate of bone turnover seems to determine aspects of bone strength that are distinct from, and are not captured by the measurement of bone mass, structure or geometry. The last 20 yr have seen considerable progress in the development and validation of serum and urine markers of bone turnover. If applied and interpreted correctly, these indices permit the specific measurement of bone formation or bone resorption without the need of going through the process of a bone biopsy. Today, there is little doubt that markers of bone turnover can be used to estimate fracture risk in both postmenopausal women and older men.1-7 However, the real domain for the use of bone turnover markers in clinical practice seems to lie in the monitoring of anti-osteoporotic therapy, treatment efficacy, and patient compliance. In contrast to changes in BMD, which occur slowly and may not be detectable before several years of treatment,8-10 biochemical markers of bone turnover change rapidly in response to therapeutic interventions.11-13 Based on these findings, and through the intense efforts of commercialization and automatization, “bone markers” are today considered useful tools that many would like to see applied in the diagnosis and management of metabolic bone disease, including osteoporosis. Thus far, however, the productive use of these markers has been confined mostly to research studies or commercial trials, which, as a rule, look at larger populations rather than the individual patient. As a matter fact, most current practice guidelines are very clear in not recommending the use of bone turnover markers in routine clinical practice. The reasons for the apprehensiveness of most independent experts and professional societies in regard to the widespread use of bone turnover markers are 3-fold: first, the variability of nearly all bone markers makes it difficult to apply the results derived from larger trials to individual patients; second, many countries lack pivotal quality control programs for bone turnover markers, including those offered by commercial laboratories-this deficit is further confounded by poor standardization of most assays; and third, the lack of valid reference ranges leaves many clinicians with uncertainties how to interpret a given result. It is in this context that the paper by Glover et al.14 in this issue of the Journal is of particular interest. Based on an international study of 637 premenopausal women, the investigators aimed to set-up reference intervals for two well-established markers of bone formation (bone alkaline phosphatase [ALP], N-terminal propeptide of type I procollagen [PINP]) and two markers of bone resorption (serum C-terminal telopeptide of type I collagen [S-CTX], urinary N-terminal telopeptide of type I collagen [U-NTX]). Premenopausal volunteers eligible to participate were women between the age of 30 and 39 yr with a BMI between 18 and 29 kg/m2 and without medical conditions known to affect bone metabolism (although some were taking an oral contraceptive pill). All samples were obtained at roughly the same time of the year (in the northern summer) and time of day (i.e., between 8:00 a.m. and 10:00 a.m.) after an overnight fast. The study is strengthened by its relatively large sample size, its reasonably well-characterized study population, and a strict study protocol within the wider context of the HORIZON-PFT trial. Although the authors paid special attention to minimize variability in their study cohorts, serum levels of bone turnover markers (PINP, S-CTX) were found to be different between study centers in the United Kingdom, France, Belgium, and the United States. For instance, women in France showed significantly higher mean levels of S-CTX relative to those living in the United Kingdom and significantly higher mean levels of PINP than women living in Belgium and the United Kingdom. The regional differences observed in this study were in part explained by variations in BMI and lifestyle factors such as smoking habits, use of the oral contraceptive pill, exercise levels, and alcohol consumption. These findings again highlight the clinically important fact that bone marker levels are affected by a number of modifiable and nonmodifiable factors. Of note, the authors clarify that the modifiable (i.e., controllable) factors explained only a small proportion of the overall variability of bone marker levels in these healthy premenopausal women. Many other factors such as environmental, genetic, or medical influences were not controlled for, and it is conceivable that, as a rule, those factors would be hard to control for in everyday clinical practice. Therefore, even with robust country-specific reference ranges available, clinicians will still have to take into account a host of more or less uncontrollable factors that affect the variability and hence validity of any bone marker measurements in their individual patients. Whereas such reference ranges are certainly helpful as guides through a particularly uneven terrain (especially for large-scale trials and research studies), they only partly eliminate the uncertainties in the interpretation of a bone marker value in an individual patient. The authors recommend that the country-specific reference intervals should be applied to the populations in which these ranges were established. That makes sense but leaves us with the question of what physicians should do in Switzerland, Germany, Australia, or in any other country that had not participated in this study. Can these reference intervals be extrapolated to other populations and ethnicities? The answer is “No,” and whereas Glover et al. suggest that countries that did not participate in their study should establish their own reference intervals locally,14 this seems a cumbersome and impractical approach. Even if we were to accept this solution, such reference ranges would need to be established in a centralized laboratory participating in external quality assurance programs and using automated assay technology to reduce sources of error (as was the case in the study by Glover et al.). However, the real world differs from specialized research centers in that clinicians will use any number of public or private laboratories for ordering their bone marker measurements, often changing between different laboratories depending on their location of practice and other variable circumstances. As with many other tests, the interpretation of bone marker levels and their changes during treatment depends on accurate technique and reproducible results. Unfortunately, interlaboratory variation has been shown to be a source of analytical variability that should not be underestimated. A proficiency testing program for a large panel of biochemical bone markers offered by >70 private and public clinical laboratories in different European countries has shown unacceptable between-laboratory variation for almost all markers tested, irrespective of the fact that only commercially available and presumably standardized assays had been used.15 Thus, in identical samples, results obtained for the same marker by the same method differed up to 7.3-fold. Based on these findings, results for most markers of bone turnover can not be compared among different laboratories without previous cross-calibration. One can of course make the assumption that the use of newer automated assays will minimize analytical variation between laboratories provided that these laboratories participate in external quality assurance programs. However, proof of the first assumption is still outstanding, and to our knowledge, very few laboratories take part in specific quality assurance programs for bone markers. Finally, all reference ranges available from the published literature14,16,17 are based on healthy premenopausal women. What about men? Assuming that preanalytical and analytical variability is comparable between women and men, interpretation of bone turnover in men is, in part, hampered by the lack of sex-specific reference ranges. To this end, further studies are needed to facilitate an accurate and broader use of bone markers in clinical practice. The reference range established by Glover et al.14 is in accordance with those established in previous studies including premenopausal women from the United Kingdom16 and from seven sites across the United States.17 Ideally, clinicians might therefore rely on these reference intervals and might translate them into clinical practice when interpreting bone marker levels. Nevertheless, we need to acknowledge that, in the real world, things are more complicated. Preanalytical and biological variability remains a key issue in the interpretation of an individual patient's results. Finally, strict quality control programs for bone turnover markers should be implemented worldwide as a first step to improve standardization of those assays most widely used in clinical practice.