Is There a Place for Bone Turnover Markers in the Management of Osteoporosis?

A personalized approach to the diagnosis and treatment of osteoporosis requires the reliable estimation of future fracture risk as this will allow us to direct treatment to the patients who are most likely to benefit from the treatment. Although BMD is a good predictor of future fracture risk, it is not perfect. Many patients suffer fragility fractures at osteopenic or even normal BMD. There is, therefore, a need to improve fracture risk prediction for the individual patient. Several fracture risk assessment tools have been developed over the last two decades; they incorporate clinical risk factors in the fracture risk assessment—this has improved the risk assessment somewhat. Other aspects of bone quality have been added to the assessment armamentarium; for example, trabecular bone score provides information about microarchitecture and has been shown to improve fracture risk prediction. Bone turnover also influences bone quality and strength by destabilizing the trabecular network and inducing cortical porosity. Bone turnover can be assessed in many ways: from invasive bone biopsies to a fasting blood sample in which biochemical markers of bone turnover can be measured. Many studies over the last three decades have investigated the ability of various bone turnover markers to predict future fracture risk. The outcomes of these studies have been variable: Some found bone turnover markers to predict future fracture risk, but many did not. The most recent of these studies is reported in this issue of the Journal of Bone and Mineral Research. Crandall and colleagues report a nested case-control analysis of the predictive value of bone turnover markers for hip fracture in women participating in the Women's Health Initiative (WHI) study. They measured s-PINP and s-CTX at inclusion in the WHI study and identified 400 women, who suffered a fragility hip fracture between August 1994 and August 2004, and 400 control subjects without a hip fracture matched for age, race/ethnicity, and date of inclusion. The average follow-up period was a little over 7 years. Baseline levels of the two markers were not different between the cases and controls, and neither of the markers predicted incident hip fractures.1 These findings are in accordance with the majority of studies investigating the predictive value of bone turnover markers for risk of future hip fractures. The most cited exception is the EPIDemiology of OSteoporosis (EPIDOS) study,2 where an association was found between s-CTX and incident hip fracture. Important differences between the study by Crandall et al. and the EPIDOS study include age differences, the duration of follow-up, and adjustments for other relevant covariates. An analysis from the Rotterdam study demonstrated that an association between urinary-free deoxypyridinoline and incident hip fracture disappeared after correction for disability.3 Why are the results of these studies and the meta-analyses of these studies4 conflicting? Bone turnover markers are different from BMD or TBS and most other aspects of bone quality. BMD and TBS are relatively stable over time and can be measured with very high precision. Bone turnover markers have a diurnal rhythm and are affected by food intake—in some individuals more than others—as well as by many other factors. In addition, analyses of individual markers have been technically challenging, resulting in imprecise measurements. However, although the technical aspects have recently improved significantly, the variability is still large compared to the variability seen with the measurement of BMD or TBS. Should we then give up on using biochemical markers of bone turnover in the clinic? Based on the available evidence, bone turnover markers do not have an important role in the assessment of long-term fracture risk in an untreated patient, but high levels of bone turnover markers in a patient have been suggested to be partly to blame for the high imminent fracture risk observed in patients with a recent fragility fracture.5 High levels of biochemical markers of bone turnover in a patient with osteoporosis should lead to further investigation of other causes of low bone mass or fracture as osteoporosis is not generally characterized by high bone turnover, except in early postmenopausal women whose bone turnover often is high. Where the responsiveness of bone turnover markers to many different factors makes them less good at predicting long-term outcomes such as fractures, it makes them better for monitoring interventions—pharmacological and to some extent nonpharmacological.6, 7 Bone resorption markers and subsequently bone formation markers respond quickly to antiresorptive therapy and remain suppressed during effective antiresorptive therapy. The response to bone-forming treatment is a rapid and pronounced increase in bone formation markers followed by an increase in bone resorption markers; the pattern of changes is slightly different with teriparatide and abaloparatide. Because the changes are prominent, they can be assessed in the individual patient despite the preanalytical and analytical variation associated with the measurements. The bone turnover markers are also useful when considering a treatment break in a patient on long-term bisphosphonate treatment. If bone turnover markers are not suppressed in a patient, then the patient is either not treatment-compliant, is not absorbing the oral bisphosphonate sufficiently well, or has another reason altogether, such as hyperthyroidism, vitamin D deficiency, malignancy, etc. This patient should not go on treatment break, but be re-evaluated. If no other reason for the nonsuppression of bone turnover can be found, another anti-osteoporosis treatment should be considered. However, if bone resorption is suppressed and other recommendations regarding treatment breaks have been explored, the patient can go on treatment break with bone turnover markers used to follow the patient during the treatment break to help decide whether to restart treatment. However, many patients with osteoporosis can be managed without the use of bone turnover markers. If the patient takes the medication as prescribed or is treated with a parentally administered treatment and does not experience fractures, there is no need to measure bone turnover. In life, it is important to pick the optimal tool for the job in front of you. In the management of patients, BMD, probably in combination with TBS and clinical risk factors, is currently the best tool for the prediction of future fracture risk. Bone turnover markers are best used to monitor the interventions we initiate against an increased risk of future fractures.8, 9