TBS iNsight to boost your DXA

The bone quality assessment technique for enhancing identification of fracture risk

Introduction

The World Health Organization defines osteoporosis as a silent disease characterized by low bone mass (bone density) and microarchitectural deterioration of bone tissue leading to increased bone fragility and elevated risk of fracture[1]. Worldwide, osteoporosis affects an estimated 200 million women and causes nearly nine million fractures annually[2, 3]. Globally, one in three women and one in five men over the age of 50 will experience a fracture due to osteoporosis[4, 5] with a subsequent decrease in quality of life and an excess mortality rate for hip fractures >20% in the first year[6]. By 2050, the worldwide incidence of hip fracture in women is projected to increase by 240%; and in men by 310%[7].

Bone densitometry (DXA: dual-energy x-ray absorptiometry) is accurate, painless and readily accessible in most communities. For these reasons, DXA has become well accepted as a standard tool for the assessment of osteoporosis. DXA utilizes x-rays of two distinct energies to provide quantitative information related to bone mineral density (BMD). However, this does not always sufficiently translate into an accurate estimate of future fracture risk.

The bone quality assessment technique for enhancing identification of fracture risk Moreover, it is now well established that BMD is not the only characteristic of bone that determines its strength and fragility and, therefore other aspects must be considered when deciding upon therapy to prevent new or further osteoporotic fractures[8]. For example, it is well known that over 50% of fractures occur in patients with BMD values that are not classified as “osteoporotic” according to the WHO classification of osteoporosis (figure 1)[9]. This observation implies that factors other than BMD influence bone strength and fracture risk. These factors include bone macro-geometry, bone mineralization, and bone turnover[9, 10]. Another key determinant of bone strength is its micro- architecture, the importance of which has been increasingly appreciated in recent years, on top of the fact it was already implied from the conceptual definition of osteoporosis[10]. This acknowledgement has led to the recognition that evaluating bone micro-architecture might significantly enhance the accuracy of bone strength evaluations and, consequently, also of fracture risk[11, 12].

TBS iNsight: The Tool to Refine Patients’ Risk Profile

TBS iNsightTM is a software tool that installs on most existing GE and Hologic DXA scanners. This simple, rapid and reproducible method estimates fracture risk based on a determination of bone texture (an index correlated to bone microarchitecture) [13, 14], in addition to risks determined by DXA based bone mineral density, clinical risk factors and FRAX. The result is expressed as a Trabecular Bone Score (TBS).

It requires no additional scan time or additional radiation exposure nor extra work for the technician. Once the standard DXA spine scan is completed, TBS results are displayed automatically. TBS iNsight enables retrospective analysis of older DXA scans (prior exams must be acquired on the same DXA unit with a valid TBS calibration).

How It Works

TBS is a texture index that evaluates pixel gray-level variations in the lumbar spine DXA image, providing an Simply stated, TBS principles are based on the fractal property of 2D projected bone microarchitecture[16]. The DXA spine scan does not have sufficient resolution to identify individual trabeculae. However, a dense trabecular microstructure projected onto a plane generates an image containing a large number of pixel-to-pixel gray-level variations of small amplitude, whereas a 2D projection of a porous trabecular structure produces an image with a low number of pixel-to-pixel gray-level variations, but of much higher amplitude. In other words, different bone microstructures will appear differently on the DXA image and that difference is captured through the TBS analysis (figure 2).

A variogram of those projected images, calculated as the sum of the squared gray-level differences between pixels at a specific distance, can estimate a 3D structure from the existing variations on the 2D projected images. TBS is derived from the experimental variograms of 2D projection images. TBS is calculated as the slope of the log-log transform of the variogram, where the slope characterizes the rate of gray-level amplitude variations. indirect yet highly correlated evaluation of trabecular A steep variogram slope with a high TBS value is associated with better bone structure, while low TBS values indicate worse bone structure.

TBS Clinical Evaluation

TBS has been used in more than 400 peer-reviewed publications worldwide and on more than 75,000 patients to address several scientific and clinical questions. TBS has repeatedly been proven to be predictive of fragility fractures (current and future) and this largely independently of BMD, clinical risk factor and the FRAX based risk estimates; and, when used in conjunction with any one of these measures, it consistently enhances their accuracy. There is also a growing body of evidence indicating that the TBS has particular advantages over BMD for specific causes of increased fracture risk, like chronic corticosteroid use, type-2 diabetes, chronic kidney disease, primary hyperparathyroidism, patients being treated with anti-aromatase, conditions where BMD readings are often misleading [17, 18].

Some of the key findings have been conveniently summarized in recent review articles published by groups of international bone experts [17, 19-23]. A short list of pivotal studies is reported in table 1. The main points are summarized below:

• TBS is lower in men and postmenopausal women with prevalent vertebral, hip or major osteoporotic fractures compared to controls.
• TBS predicts incident major osteoporotic fractures, spine and hip fractures in women and men independently of both lumbar spine BMD measurements and clinical risk factors; TBS is therefore complementary to these existing approaches. The greatest utility lies in individuals whose BMD levels are in the osteopenic range.
• TBS can be used as an adjustment parameter of the FRAX tool to better predict osteoporotic fractures in conjunction with other clinical risk factors [figure 5]. Added to the FRAX, the TBS’s greatest utility lies in individuals whose BMD levels are close to an intervention threshold (up to 25% of the patients will then be impacted).
• From a meta-analysis including 14 prospective cohorts [24], TBS thresholds have been evaluated based on a tertile approach. In the high-risk tertile (TBS < 1.23), gradient of risk for major osteoporotic fracture was more than two times greater than in the low-risk tertile (TBS > 1.31) [figure 4].
• TBS can be used as an aid in the diagnosis of osteoporosis and other medical conditions leading to altered trabecular bone microarchitecture, and ultimately in the assessment of fracture risk. Diseases of interest include diabetes, hyperparathyroidism, HIV, chronic kidney disease or patients under glucocorticoid use or under anti-aromatase treatment [20].
• The short-term reproducibility of TBS measurements has been reported in several studies with values ranging from 1.1% - 2.1% C.V [19];
• Although BMD is more reactive (in amplitude) to the different treatments affecting bone metabolism, the differential effect of these different pharmaceuticals on TBS may have its usefulness in routine clinical practice. As such, TBS may assist physicians in monitoring the response to treatments over time [table “” not found /]
  .
• Unlike BMD, TBS results have been demonstrated to be minimally affected by the presence of osteophytes – a common artifact in late postmenopausal patients and those presenting with osteoarthritis [25];
• TBS has been endorsed by many local, national and international medical societies [table “” not found /]
  .

Possible Interpretation of TBS values in overall patient management

The TBS report is generated simultaneously with the standard DXA spine printout. The report (figure 3) shows an overall Trabecular Bone Score, displays the TBS mapping of the spine, and provides age-matched reference values.

TBS can be easily combined with BMD T-score as the interpretation table shows in figure 4. This interpretation table is adapted from a meta-analysis [24] and the Manitoba study[26] and provides a class of fracture risk for major osteoporotic fracture which depends on both WHO T-score zone for BMD (normal, osteopenic and osteoporotic) and on TBS thresholds. For example: an osteopenic woman with a -2.2 T-score at the lumbar spine falls into a risk class of major osteoporotic fracture of about 5 to 7 per 1000 women per year. Adding the patient’s TBS value (1.180) to the picture, moves her into a superior risk category corresponding to 10 to 14 fractures per 1000 women per year. That is to say, this woman’s combined fracture risk is similar to the fracture risk of an osteoporotic woman. This example demonstrates how TBS can be used to better evaluate a patient’s risk of fracture and then to improve the overall patient care management.

Improve fracture prediction with TBS-adjusted FRAX

TBS can be used easily as a FRAX modifier (figure 5). As recommended by the ISCD[12] and IOF/ESCEO[11], TBS can be used in association with FRAX and BMD to adjust FRAX- probability of fracture in postmenopausal women and older men [12]. The FRAX tool is based on individual patient models that integrate the risks associated with clinical risk factors as well as BMD at the femoral neck.

Models for adjusting fracture probability from FRAX to account for TBS were derived in large population-based cohorts [46] and cross-validated in a meta-analysis including 17,809 men and women from 14 prospective population- based cohorts [24]. Authors found that for both hip fracture and major osteoporotic fracture, incorporation of the TBS-adjustment factor resulted in an improvement in the gradient of risk. Other independent studies reported an improved fracture prediction using TBS-adjusted FRAX in primary or secondary osteoporosis [47- 49].

Use of TBS to Monitor Treatment: Review of Selected Studies

The TBS parameter, as being influenced by trabecular pattern, might also be influenced by treatments known to impact bone microarchitecture. TBS has been used in various pharmaceutical trials designed to evaluate the effect of osteoporosis treatments, either antiresorptive (slow down bone destruction) or anabolic agents (aimed at rebuilding bone). Bisphosphonates (alendronate, zoledronate, etc.) and denosumab belong to the antiresorptive category, while teriparatide is classified as an anabolic agent. These studies, summarized in table 3, compared the effect of drugs either against placebo or against another reference drug.

Pooled results are represented in figure 6.

Interestingly, the various efficacious therapies for osteoporosis differ in the extent to which they influence the TBS, with bisphosphonates exerting very little effect, but other drugs like PTH / PTH analog generally increasing TBS in the range of one to two percent per year. These findings seem to be consistent with the mechanism of action of the molecules. Indeed, one would not expect to see an improvement of the micro-structure with a bisphosphonate (and so TBS) while the degree of mineralization would increase and thus also BMD. These primary studies start to show the interest of evaluating both BMD and TBS during treatment monitoring.

Taken together, these studies suggest that TBS tends to increase with treatments that increase BMD. The magnitude of TBS increase is usually less marked than BMD changes. In contrast, the magnitude of the decrease in TBS without treatment is very similar to that of BMD. It seems clinically relevant to consider that an increase of surrogate markers of both bone quantity (BMD) and quality (TBS) would be reassuring to monitor effects of treatments.

Summary

Trabecular bone score (TBS) is a grey-level textural measurement derived from lumbar spine dual-energy X-ray absorptiometry (DXA) images. It is related to bone microarchitecture that provides skeletal information complementary to that obtained from standard bone mineral density (BMD) measurement.

This summary paper documents a unique way to assess bone texture, a surrogate of bone microarchitecture and subsequently bone strength that not only predicts future fracture risk: it does so independent of BMD, clinical risk factors and the FRAX tool. It also enhances the accuracy of these tools when added as a supplementary test. Moreover, it demonstrates diagnostic accuracy for both primary and secondary osteoporosis and in both females and males, and appears sensitive to change over time that are the result either of effective treatment (with TBS increasing) or continued bone loss in the absence of effective treatment (with TBS decreasing). In some scenarios — for example in patients with type 2 diabetes or disorders associated with increased extraneous calcification around the spine, like degenerative spine disease or ankylosing spondylitis — it almost outperforms even the gold standard diagnostic measure for osteoporosis: DXA measured BMD.

Practically, TBS as an adjustment parameter of FRAX enables physicians to benefit from a more accurate evaluation of fracture risk with no change in the existing workflow.

Using FRAX Adjusted for TBS allows physicians to:

• Integrate TBS easily in daily clinical practice
• Enhance fracture predictability using FRAX
• Refine individual fracture risk assessment
• Tighten selection of patients in need of therapeutic treatment.

TBS iNsight is therefore a useful tool to enhance fracture risk prediction in clinical settings in conjunction with BMD and clinical risk factors.

References