Abstract
Osteoporosis is a common comorbidity of all major rheumatic diseases, and manifests itself both systemically and locally. Systemic bone loss manifests because of several factors, primarily inflammation, immobility, and commonly used medical treatment for rheumatic diseases. Local bone loss manifests as periarticular demineralization and bone erosion due to local release of inflammatory agents and cytokines, which promote bone resorption. All these factors contribute to the phenomenon of arthritis-associated osteoporosis. This review summarized the currently available and used methods that play a role in the diagnosis and monitoring of osteoporosis and in the detection of osteoporotic fractures.
History of imaging in osteoporosis in the past and present
Ultrasound (US) was the first imaging method that was used to assess the mechanical properties of cortical bone in the 1960–70’s, and this was later extended to include the evaluation of calcaneal trabecular bone. Quantitative US (QUS) measurement was indeed an attractive modality for evaluating bone, because of its relatively low cost and lack of radiation. The velocity and attenuation of the US wave depends on the physical properties of the medium, and it can therefore be used to characterize the medium through which it travels .
At approximately the same time (1960’s), bone mineral assessment using single-photon isotope of the peripheral skeleton (radius and calcaneus) and dual-photon absorptiometry (SPA and DPA, respectively) of the spine, hip, or entire patient was introduced. The method used either a iodine-125 radiation source (SPA) and gadolinium-153 photon source (DPA), and was substituted with the dual-energy X-ray absorptiometry (DXA) scanner in the 1980’s, which was based on dual-energy X-ray source and not a radioisotope. DXA was similar to DPA in terms of both method and performance with a correlation coefficient of 0.99 . In general, bone mineral density (BMD) measured using DXA was lower than those measured by DPA, which was due to a more accurate assessment of the bone area .
Quantitative computed tomography (QCT) quantifying BMD in the spine, proximal femur, forearm, or tibia was introduced in the 1970’s as one of the first quantitative methods for bone imaging. The main advantages of QCT over DPA were its capability for precise three-dimensional (3D) anatomic localization providing direct density measurement, and its capability for spatial separation of highly responsive cancellous bone from less responsive compact bone . However, as DXA uses considerably lower radiation doses, it was this method that was eventually used to define osteoporosis by the World Health Organization (WHO) criteria .
The association between BMD and fracture risk is well known from clinical trials; yet, we also know that BMD measurement is insufficient for fracture risk prediction. Indeed, bone strength is determined by both bone mass and quality. As DXA only evaluates bone density, newer quantification methods and modalities have been introduced, which enable the assessment of bone quality.
Magnetic resonance imaging (MRI) and magnetic resonance spectroscopy (MRS) are newer methods, which allow an assessment of bone architecture and bone marrow composition. By the separation of water and fat signal in the vertebrae or hip, the amount of bone marrow fat (BMF) can also be evaluated. High-resolution peripheral QCT (HRpQCT) allows the evaluation of both trabecular and cortical bone architecture. To better assess bone quality, the trabecular bone score (TBS) is another novel imaging modality used to describe skeletal microarchitecture based on information derived from the DXA image of the lumbar spine.
An important factor, which has limited the introduction of new diagnostic modalities, is the need for comparison with a gold standard. In the case of osteoporosis, the DXA scan was and remains the gold standard imaging method and the basis of the definition of disease and requirement for certain treatment decisions in many countries. The main challenge is that DXA does not predict fracture risk alone. Several studies, particularly in elderly women, have shown that the risk of fracture roughly doubles for each standard deviation (SD) reduction in BMD . This is contradiction with the fact that the annual incidence of hip fracture increases approximately 30-fold between the ages of 50 and 90 years . However, based simply on the known relationship between BMD and fracture risk and the loss of bone with age, we would expect that the risk of hip fracture would rise only fourfold. Therefore, the increase in risk with age cannot be explained on the basis of BMD alone.
Patients with poor bone quality may thus exhibit normal bone mass, and may not be diagnosed as having an increased risk of osteoporotic fracture until the day they actually experience one. The aim in osteoporotic research in recent years has thus been to identify new modalities that alone or in combination with DXA and known risk factors can predict fracture risk better, and identify patients with poor bone quality and normal or osteopenic bone mass before they experience a fracture.
Osteoporosis is an important comorbidity of many chronic diseases, and inflammatory rheumatic disease is one of them. This is due in part to the effects of inflammation, a process common to these diseases, which leads to increased osteoclast activation and subsequent bone resorption mediated by proinflammatory cytokines . Uncontrolled disease activity has a detrimental effect on bone density. Furthermore, pain, fatigue, and functional dysfunction, which frequently accompany rheumatic diseases, all lead to physical inactivity, further increasing risk of osteoporosis. Although appropriate treatment can negate this effect, unfortunately a number of agents that are used commonly to treat rheumatic conditions have adjunctive effects on bone loss in rheumatic patients. These factors have all contributed to the coining of the term “rheumatoid-arthritis associated osteoporosis” . Indeed, a large number of epidemiologic studies have demonstrated that not only patients with rheumatoid arthritis (RA) but also those with systemic lupus erythematosus (SLE), ankylosing spondylitis (AS), psoriatic arthritis (PsA), myositis, and systemic sclerosis (SS) are at an increased risk of osteoporosis and fragility fractures. Thus, it appears that the phenomenon of “arthritis-associated osteoporosis” is common to the entire spectrum of rheumatic and musculoskeletal diseases (RMDs) and should be more appropriately termed “RMD-associated osteoporosis.”
Use of imaging modalities in the diagnosis of osteoporosis
Dual-energy X-ray absorptiometry
Measuring bone mass by DXA is based on the emission of X-rays of two distinct energies. The subsequent computer analysis eliminates the soft tissue component and yields the BMD result. Using dual-energy sources allows the differentiation of soft tissue from bone tissue. BMD measurement by DXA is considered the gold standard for diagnosing osteoporosis and is the basis of the WHO definition of osteoporosis . The region of interest (ROI) when performing a DXA of the lumbar spine is from L1 to L4 or sections thereof. The patient is lying in the supine position, centered on the DXA scanner, with hips and knees flexed on a position block. This position straightens the lumbar lordosis for reproducible results. The X-ray beam enters the body from posterior and travels in the anterior direction. The DXA image must show equal amounts of soft tissue on both sides. The superior margin of both iliac crests, the middle of the T12 vertebra, and the middle of the L5 vertebra must be visible on the acquired image.
Several ROI are present for the DXA of the hip: femoral neck, trochanter, intertrochanteric, and total hip. Ward’s triangle is not a true anatomic area but an area generated by the DXA scan, which has the lowest BMD in the femoral head. The measurement of BMD in Ward’s triangle should not be used to diagnose osteoporosis, because it leads to overestimation of osteoporosis . One should use the BMD measurement taken at either the femoral neck, the trochanter, or total hip BMD. The patient is in the supine position with the hip rotated to a degree where the lesser trochanter appears small or not visible, and the DXA image shows the femoral shaft aligned with the long axis of the scanner.
Measurement of the forearm region is used when the standard spine or hip region is not measurable or cannot be interpreted. Measurement of the forearm region is not included in the diagnosis of osteoporosis, but can be a useful tool to evaluate bone over time in the clinical setting . Overall, the main advantages of DXA include its general availability, low radiation exposure, and low cost ( Table 1 ). It is however a measurement that calculates the bone mineral content in grams and the 2D projected area in cm 2 ; thus, the BMD of the bone(s) is measured in terms of g/cm 2 . It therefore measures only density/area and not the density/volume as seen with QCT. BMD measured by DXA is influenced by bone size and will thus overestimate fracture risk in individuals with small body frame, who will have lower areal BMD than normal-sized individuals, and is also influenced by degenerative changes, primarily those affecting the spine. Such patients have increased areal density, particularly in areas of bony growth, which will suggest a lower fracture risk than is actually present. In addition, all structures overlying the spine, such as aortic calcifications, or morphologic abnormalities, such as after laminectomy at the spine, will affect BMD measurements .
| Method | Availability | Radiation | Bone density | Bone quality | Low cost | Estimation of fracture risk |
|---|---|---|---|---|---|---|
| DXA | +++ | + | +++ | + | +++ | +++ |
| QCT | ++ | +++ | +++ | +++ | + | +++ |
| HRpQCT | + | ++ | +++ | +++ | ++ | +++ |
| QUS | ++ | − | +++ | ++ | +++ | ++ |
| MRI/MRS | + | − | − | +++ | + | ++ |
The working group of the WHO proposed to define osteoporosis on the basis of the difference between the measured BMD and the mean value of young adults, expressed in SD for a normal population of the same gender and ethnicity. This derivative is known as the T-score ( Fig. 1 ). On the basis of the DXA result, postmenopausal women and men aged >50 years are stratified into one of the following categories according to T-score result. Normal BMD: T-score > −1; osteopenia: T-score >−2.5 and below or equals T-score −1; and osteoporosis: T-score <−2.5. The definition was based on the calculation of the prevalence of osteoporosis, aiming at a cutoff value of 30%. Importantly, this definition does not include premenopausal women, young men, and children. In addition to the T-scores, DXA reports also provide Z-scores, which are calculated similar to the T-score, except that the patient’s BMD is compared with an age-matched (and race and gender matched) mean, and the result is expressed as an SD score. The Z-score is no longer included in the definition of osteoporosis, but in premenopausal women, young men, and children, a low Z-score (<−2.0) may indicate that bone density is lower than expected and should trigger a search for an underlying cause. It should be noted that even in the absence of T-score >−2.5, the presence of one or more low-impact fragility fractures is considered as a sign of severe osteoporosis . Given the limitations of DXA BMD measurements, the WHO recently introduced the Fracture Risk Assessment Tool (FRAX) to better evaluate fracture risk of patients. On the basis of clinical risk factors and DXA BMD of the femoral neck, the FRAX tool can be used to calculate the 10-year probability of hip fracture and major osteoporotic fractures (spine, forearm, hip, or shoulder fracture) .
Quantitative ultrasound
QUS evaluates primarily bone density with no radiation and because it is also inexpensive and generally available, it has been tempting to introduce this modality in the diagnosis and evaluation of osteoporosis. Two main variables can be measured by QUS devices, which derive from velocity or attenuation of the US waves through the bone tissue. The variables reflecting US velocity inside the bone, expressed as meter per second, are known as speed of sound (SoS), which is a pure parameter of velocity independent of US wave attenuation. SoS is a variable usually measured by QUS methods applied to the heel, radius, tibia, and patella. The most common variable reflecting US attenuation through bone is known as broadband ultrasound attenuation (BUA), a measure of the frequency dependence of the attenuation of the signal, which is expressed as dB/MHz and is commonly assessed by calcaneal QUS devices .
A recent systemic literature review comparing QUS against DXA as a diagnostic tool revealed large variations in specificity (28–71%) and sensitivity (65–93%) depending on the study population and DXA region . Because availability of DXA scanners in some regions of the world is limited, QUS was considered as an attractive alternative for osteoporosis screening. However, one finds large heterogeneity between studies and uncertainty regarding cutoff, device, and measured variable and misclassification rates of osteoporosis ranging from 0% to 12.4%, when compared with DXA . Although QUS has been shown to identify individuals with fragility fractures and to predict fracture risk, it has not been established as a tool to diagnose OP as application of the DXA cutoffs to the QUS measurement was found to underestimate the true prevalence of osteoporosis .
Currently, more sophisticated QUS indices derived from the two basic measurements are available, such as amplitude-dependent SoS (AD-SoS), bone transmission time (BTT), stiffness index (SI), quantitative ultrasound index (QUI), which evaluate bone quality and estimated BMD (eBMD). These new composite parameters may be useful in the determination of subjects with low bone health status .
Quantitative computed tomography
QCT is based on X-ray and measures the absorption coefficient in specific tissue compared to that of water. The result is expressed in Hounsfield units (HU). QCT may in principle be performed on any CT scanner, with the use of a calibration phantom and dedicated software. Although different phantoms exist, they cannot be interchanged without cross-calibration. The patient is scanned in the supine position, laying on the phantom. The calibration phantom is used to convert the obtained HU into a BMD measure expressed in mg hydroxyapatite/ml. QCT measures 3D volumetric density, as opposed to being a projected measure as DXA–BMD. The ROI can be separated from degenerative changes in the spine and may serve as a problem-solving tool if DXA is difficult to interpret, for example, in very small or large patients, as the absolute BMD result is independent of body size. It provides purely trabecular bone measurements, which are more sensitive to monitoring changes with disease and therapy than DXA. However, as radiation dose is high (factor 50–100), it is not recommendable as a substitution to DXA unless it is expected to provide superior information to that of DXA. QCT has also been used as a tool for better in vivo quantification the bone microarchitecture. It is difficult to verify the data obtained in in vitro studies in clinical practice, because of the need of a high radiation dose. Compared with the 0.01–0.05 mSv effective dose associated with DXA in adult patients and 0.06–0.3 mSv delivered through 2D QCT of the lumbar spine, high-resolution multidetector CT (MDCT) used to examine vertebral microstructure provides an effective dose of approximately 3 mSv, equivalent to approximately 1.5 years of natural background radiation . Thus, the hope of integrating knowledge about bone quality into the prediction of fracture risk is not clinically applicable using QCT bone quality parameters presently.
QCT of the spine
Single-slice QCT of the spine evaluates 3–4 vertebrae in a row (T12–L4), using 8- to 10-mm-thick slices of the midvertebral sections parallel to the endplate. The radiation dose is higher than that used in DXA and the precision is lower. Volumetric QCT (vQCT) of the spine uses 2 complete vertebrae in the region Th12–L3 (often L1–L2) for analysis. vQCT has a higher radiation exposure dose than single-slice QCT, but spatial resolution is better due to thinner slices (1–3 mm). Using MDCT, BMD precision is 1–2.5% high and comparable to DXA . Compared to the hip, the spine has a larger proportion of the trabecular bone. Because the metabolic activity of the trabecular bone is higher and because QCT separates the trabecular and cortical bone department, a larger decline in QCT–BMD spine is observed than that observed in DXA–BMD spine, which measures integral bone. Applying the WHO defined T-scores of osteoporosis results in an overestimation of the disease. Therefore, the American College of Radiology has suggested guidelines for evaluating spine QCT–BMD results: BMD between 120 and 80 mg/ml is defined as osteopenia and values <80 mg/ml as osteoporosis .
QCT of the hip
Single-slice technique cannot be used in the hip region, because of the more complex anatomy. QCT of the hip is performed from a few centimeters above the femoral head to a few centimeters below the lesser trochanter. A volumetric image is processed and BMD of the femoral neck, total hip, and trochanteric regions are derived using a 2D projection. Statistically significant differences in discriminatory power were not found in BMD of the spine obtained with vQCT or DXA; however, integral BMD of the spine measured by vQCT and DXA tended to show stronger associations with fracture status . Similar to the QCT of the spine, the QCT of the hip is not recommendable as a substitution to DXA unless it is expected to provide superior information to that of DXA. Volumetric BMD of the spine and hip has been shown to correlate to those measured from nondedicated standard MDCT scans. When using a well-defined, regional specific equation, correlation coefficients of r = 0.98 and 0.99 have been reported for the spine and hip, respectively, making the strategy promising in increasing the frequency of diagnosis . QCT–DXA of the hip (CTXA of the hip) shows a very high correlation to the DXA–BMD, and comparing results with a reference database, the WHO definition of osteoporosis may be applied (T-score < −2.5) . QCT may give additional information about hip geometry by using finite element analysis (FEA), a mathematical technique to evaluate the strength of complex structures. In vitro testing of the FEA result demonstrates a high degree of correlation, and integrating density and geometry analysis has been shown to be a strong predictor of bone strength and fracture risk prediction in clinical studies .
High-resolution peripheral quantitative computed tomography
Providing information about BMD and trabecular and cortical bone architecture, HRpQCT is substantially characterized by higher signal-to-noise ratio and spatial resolution than MDCT and MR. An effective radiation dose substantially lower than that of whole-body MDCT and a scan time of approximately 3 min for the tibia or femur are among the favorable properties of HRpQCT. As implied by its name, however, it is limited to the peripheral skeleton and therefore may not provide direct information about the lumbar spine or the hip – common sites for osteoporotic fragility fractures. A five-cylinder hydroxyapatite calibration phantom is used to generate volumetric BMD separately for cortical and trabecular bone compartments similar to central QCT. HRpQCT was shown to be reproducible and allows the calculation of both morphometric indexes analogous to classic histomorphometry (bone volume/tissue volume, trabecular thickness, etc.) as well as biomechanical properties (e.g., stiffness and elastic modulus; Fig. 2 ) .
Magnetic resonance imaging and magnetic resonance spectroscopy
Trabecular architecture can also be evaluated with MRI. MRI-derived trabecular structural measures have been shown to correlate with histology, micro-CT, and biomechanical strength . Although lack of radiation makes MRI seem like an ideal imaging method to assess bone quality, its spatial resolution in the range of trabecular dimensions results in substantial partial volume effects, and long acquisition times make imaging susceptible to motion artifacts. Recently, ultrashort-echo-time (UTE) imaging techniques for quantification of water content of the cortical bone to assess bone quality were developed . MRS allows an assessment of bone architecture and bone marrow composition. By the separation of water and fat signal in the vertebrae or hip, the amount of BMF is quantified and expressed as a percentage. BMF is negatively correlated to BMD, and therefore, it has been speculated whether MRI could be used to diagnose osteoporosis in situations where MRI has been obtained for other purposes (e.g., due to back pain). Expression of an M-value has been suggested for comparing the result with a reference database equivalent to T-score. Evaluating BMF was found to correlate negatively with BMD evaluated by DXA or QCT . None of the clinical perspective studies have evaluated the method or compared the results against fracture risk.
Use of imaging modalities in the diagnosis of osteoporosis
Dual-energy X-ray absorptiometry
Measuring bone mass by DXA is based on the emission of X-rays of two distinct energies. The subsequent computer analysis eliminates the soft tissue component and yields the BMD result. Using dual-energy sources allows the differentiation of soft tissue from bone tissue. BMD measurement by DXA is considered the gold standard for diagnosing osteoporosis and is the basis of the WHO definition of osteoporosis . The region of interest (ROI) when performing a DXA of the lumbar spine is from L1 to L4 or sections thereof. The patient is lying in the supine position, centered on the DXA scanner, with hips and knees flexed on a position block. This position straightens the lumbar lordosis for reproducible results. The X-ray beam enters the body from posterior and travels in the anterior direction. The DXA image must show equal amounts of soft tissue on both sides. The superior margin of both iliac crests, the middle of the T12 vertebra, and the middle of the L5 vertebra must be visible on the acquired image.
Several ROI are present for the DXA of the hip: femoral neck, trochanter, intertrochanteric, and total hip. Ward’s triangle is not a true anatomic area but an area generated by the DXA scan, which has the lowest BMD in the femoral head. The measurement of BMD in Ward’s triangle should not be used to diagnose osteoporosis, because it leads to overestimation of osteoporosis . One should use the BMD measurement taken at either the femoral neck, the trochanter, or total hip BMD. The patient is in the supine position with the hip rotated to a degree where the lesser trochanter appears small or not visible, and the DXA image shows the femoral shaft aligned with the long axis of the scanner.
Measurement of the forearm region is used when the standard spine or hip region is not measurable or cannot be interpreted. Measurement of the forearm region is not included in the diagnosis of osteoporosis, but can be a useful tool to evaluate bone over time in the clinical setting . Overall, the main advantages of DXA include its general availability, low radiation exposure, and low cost ( Table 1 ). It is however a measurement that calculates the bone mineral content in grams and the 2D projected area in cm 2 ; thus, the BMD of the bone(s) is measured in terms of g/cm 2 . It therefore measures only density/area and not the density/volume as seen with QCT. BMD measured by DXA is influenced by bone size and will thus overestimate fracture risk in individuals with small body frame, who will have lower areal BMD than normal-sized individuals, and is also influenced by degenerative changes, primarily those affecting the spine. Such patients have increased areal density, particularly in areas of bony growth, which will suggest a lower fracture risk than is actually present. In addition, all structures overlying the spine, such as aortic calcifications, or morphologic abnormalities, such as after laminectomy at the spine, will affect BMD measurements .
| Method | Availability | Radiation | Bone density | Bone quality | Low cost | Estimation of fracture risk |
|---|---|---|---|---|---|---|
| DXA | +++ | + | +++ | + | +++ | +++ |
| QCT | ++ | +++ | +++ | +++ | + | +++ |
| HRpQCT | + | ++ | +++ | +++ | ++ | +++ |
| QUS | ++ | − | +++ | ++ | +++ | ++ |
| MRI/MRS | + | − | − | +++ | + | ++ |
The working group of the WHO proposed to define osteoporosis on the basis of the difference between the measured BMD and the mean value of young adults, expressed in SD for a normal population of the same gender and ethnicity. This derivative is known as the T-score ( Fig. 1 ). On the basis of the DXA result, postmenopausal women and men aged >50 years are stratified into one of the following categories according to T-score result. Normal BMD: T-score > −1; osteopenia: T-score >−2.5 and below or equals T-score −1; and osteoporosis: T-score <−2.5. The definition was based on the calculation of the prevalence of osteoporosis, aiming at a cutoff value of 30%. Importantly, this definition does not include premenopausal women, young men, and children. In addition to the T-scores, DXA reports also provide Z-scores, which are calculated similar to the T-score, except that the patient’s BMD is compared with an age-matched (and race and gender matched) mean, and the result is expressed as an SD score. The Z-score is no longer included in the definition of osteoporosis, but in premenopausal women, young men, and children, a low Z-score (<−2.0) may indicate that bone density is lower than expected and should trigger a search for an underlying cause. It should be noted that even in the absence of T-score >−2.5, the presence of one or more low-impact fragility fractures is considered as a sign of severe osteoporosis . Given the limitations of DXA BMD measurements, the WHO recently introduced the Fracture Risk Assessment Tool (FRAX) to better evaluate fracture risk of patients. On the basis of clinical risk factors and DXA BMD of the femoral neck, the FRAX tool can be used to calculate the 10-year probability of hip fracture and major osteoporotic fractures (spine, forearm, hip, or shoulder fracture) .
Related posts:
Imaging in rheumatic and musculoskeletal conditions
Imaging in the diagnosis and management of axial spondyloarthritis
Imaging of osteoarthritis (OA): What is new?
Imaging of vasculitis: State of the art
Musculoskeletal interventional procedures: With or without imaging guidance?
Imaging of regional pain syndromes; from syndromes to conditions using imaging?
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
