State of the Art Imaging of Osteoporosis

Michelle Chen, BA,* Maria Gerges, BS,*,** William Y. Raynor, MD,†,z Peter Sang Uk Park, BA,†,x
Edward Nguyen, MD,* David H. Chan, MD,* and Ali Gholamrezanezhad, MD*

Osteoporosis is a common disease, particularly prevalent in geriatric populations, which

causes significant worldwide morbidity due to increased bone fragility and fracture risk.
Currently, the gold-standard modality for diagnosis and evaluation of osteoporosis progres-
sion and treatment relies on dual-energy x-ray absorptiometry (DXA), which measures bone
mineral density (BMD) and calculates a score based upon standard deviation of measured
BMD from the mean. However, other imaging modalities can also be used to evaluate oste-
oporosis. Here, we review historical as well as current research into development of new
imaging modalities that can provide more nuanced or opportunistic analyses of bone qual-
ity, turnover, and density that can be helpful in triaging severity and determining treatment
success in osteoporosis. We discuss the use of opportunistic computed tomography (CT)
scans, as well as the use of quantitative CT to help determine fracture risk and perform
more detailed bone quality analysis than would be allowed by DXA . Within magnetic reso-
nance imaging (MRI), new developments include the use of advanced MRI techniques such
as quantitative susceptibility mapping (QSM), magnetic resonance spectroscopy, and
chemical shift encoding-based water-fat MRI (CSE-MRI) to enable clinicians improved
assessment of nonmineralized bone compartments as well as a way to longitudinally assess
bone quality without the repeated exposure to ionizing radiation. Within ultrasound, devel-
opment of quantitative ultrasound shows promise particularly in future low-cost, broadly
available screening tools. We focus primarily on historical and recent developments within
radiotracer use as applicable to osteoporosis, particularly in the use of hybrid methods such
as NaF-PET/CT, wherein patients with osteoporosis show reduced uptake of radiotracers
such as NaF. Use of radiotracers may provide clinicians with even earlier detection windows
for osteoporosis than would traditional biomarkers. Given the metabolic nature of this disease,
current investigation into the role molecular imaging can play in the prediction of this disease
as well as in replacing invasive diagnostic procedures shows particular promise.
Semin Nucl Med 54:415-426 © 2023 Elsevier Inc. All rights reserved.

Introduction decreased bone mass and increased bone fragility. It causes

an average of 8.9 million fractures worldwide on a yearly

O steoporosis, a prevalent disease especially prominent in

aging populations, is a progressive disorder of
basis.1 The risk of a lifetime osteoporotic fracture is 40%-
50% in women and 13%-22% in men.2
Osteoporosis can be divided mostly into primary osteopo-
*Department of Radiology, Keck School of Medicine, University of Southern rosis and secondary osteoporosis. Primary osteoporosis is
California, Los Angeles, CA. attributed to factors such as older age, female sex, Asian or
y
Department of Radiology, Hospital of the University of Pennsylvania, White race, a low BMI, early menopause, or a family history
Philadelphia, PA.
z
Department of Radiology, Rutgers Robert Wood Johnson Medical School,
of fractures caused by osteoporosis.3 In contrast, secondary
New Brunswick, NJ. osteoporosis often develops as a result of a chronic disease,
x
Perelman School of Medicine at the University of Pennsylvania, often endocrine or metabolic in nature, such as hypercortiso-
Philadelphia, PA. lism, hyperparathyroidism, anorexia, thalassemia, and hypo-
**Charles E. Schmidt College of Medicine, Florida Atlantic University, Boca gonadism.4-7 Additional secondary causes of osteoporosis
Raton, FL.
Address reprint requests to: Department of Radiology, Keck School of
include prolonged treatment with glucocorticoids, diabetes,
Medicine, University of Southern California, 1500 San Pablo St, Los and rheumatoid arthritis.8 The pathophysiology of osteopo-
Angeles, CA 90033. E-mail: [email protected] rosis is thus multifactorial, including such causes as estrogen

https://doi.org/10.1053/j.semnuclmed.2023.10.008 415
0001-2998/© 2023 Elsevier Inc. All rights reserved.
416 M. Chen et al.

deficiency, imbalance of bone remodeling, and abnormal Conventional Radiography

cytokine production.9
As basic radiographs still play a role in osteoporosis diagno-
Some of the risk factors associated with osteoporotic frac-
sis, management, and treatment, primarily through the diag-
tures include low physical activity, low levels or low intake
nosis of vertebral fractures, their role in screening warrants
of vitamin D and calcium, family history of fractures, low
brief discussion.
and high body mass index, and cigarette smoking.10 The risk
Within basic radiography, a semiquantitative fracture
of osteoporotic-related fractures is higher among women,
assessment proposed by Genant et al.20 classifies the severity
with older women at increased due to a decline in estrogen
of a fracture by visually estimating the extent of vertebral
levels. Therefore, it is encouraged that postmenopausal
height reduction and morphological change, whereby
women undergo routine assessments to monitor signs of
approximate degree of estimated reduction in anterior, mid-
osteoporosis.11
dle, and/or posterior height and area determines the assign-
ment of a scoring system, from borderline to severely
deformed. From this assessment, radiographers would be
able to derive a “spinal fracture index” (SFI) from the sum of
Epidemiology and Screening all grades from each individual vertebrae divided by the
Guidelines number of the evaluated vertebra. These assessment scores
and SFI are particularly useful in continuous assessment of
The U.S. Preventive Services Task Force (USPSTF), National
fractures and in relation to follow-up imaging. However,
Osteoporosis Foundation (NOF), and the American Associa-
DXA has largely replaced this modality due to its minimal
tion of Clinical Endocrinologists (AACE) propose that osteo-
radiation exposure, high-speed image acquisition, and assess-
porosis screening in women should begin at the age of 65
ment of bone mineral density, which has become the gold-
with bone density measurements.12 A10-year prospective
standard for patient assessment.
population study showed that women aged 60 to 70 years
Newer utilizations of conventional radiography include
and above 70 years old experience greater losses in BMD
the use of machine-learning algorithms to obtain a dimen-
when compared to women aged 50 to 60 years.13 Clinical tri-
sionless parameter, termed bone structure value (BSV),
als have shown that early screening for bone density pro-
which was shown by Dimai et al.21 to be highly correlated to
vided moderate benefit in preventing fractures in
DXA-derived BMD.
postmenopausal women.14
In men, osteoporosis screening also begins at 65 years or
older if they have any risk factors including anorexia, smok-
ing, alcohol use disorder or habitual drinking, having a dis-
ease associated with an increased risk of osteoporosis, or
DXA
prolonged use of specific drugs.15 Among men with osteopo- Within the field of medical imaging, central dual-energy x-
rosis, 30%-60% of cases are due to secondary causes. Within ray absorptiometry (DXA) has been established by the World
the female population, more than 50% of cases in perimen- Health Organization as the gold standard for measuring bone
opausal women are due to hypoestrogenism, use of glucocor- mineral density and diagnosing osteoporosis in postmeno-
ticoid, excess thyroid hormone, and anticonvulsant therapy. pausal women.22 Current DXA systems utilize X-ray beams
The assessment of osteoporosis thus often begins with deter- of two different photon energies allowing for the separation
mining the cause of secondary osteoporosis, usually through of attenuation values contributed by mineralized bone and
a panel of routine laboratory tests.6 soft tissue components within a skeletal site of interest,
Using central dual-energy x-ray absorptiometry (DXA) to allowing for the measurement of areal bone mineral density
diagnose osteoporosis is considered the gold standard due its (BMD) (g/cm2) which is defined as bone mineral content
ability to predict fracture risk.16 Screening frequency mostly (BMC) (g) divided by the area of measured bone (cm2).16
depends on the following factors: the patient’s age, prior When performed for the diagnosis of osteoporosis, DXA
DXA scan results, and individual risk of fracture.12 A repeat scans are applied at the lumbar spine or the proximal femur
scan within 1-2 years of starting treatment is recommended in adult patients although BMD may be measured at periph-
for those starting treatment.17 According to the National eral sites such as the distal 1/3rd radius in certain clinical sce-
Institute of Health (NIH), BMD should be measured every narios, including the evaluation of patients with
two years to provide a baseline, and more frequent scans per- hyperparathyroidism, children, or obesity.23
formed as needed for individual patients.18 The USPSTF has BMD measurements are reported for postmenopausal
suggested that accurately measuring changes in BMD women and men of age 50 years or older by use of T-score,
requires repeat screening only after 2 years have elapsed.11 defined as the number of standard deviations between the
In 2013, new guidelines were issued regarding screening BMD of the patient being imaged and the mean BMD, repre-
repeats by the Oregon Health Evidence Review Commission, senting peak bone mass, of a young, healthy adult.24 Follow-
where in women with normal BMD results, screening should ing the diagnostic criteria recommended by the World
be repeated at most every 15 years. In women with moderate Health Organization, a normal BMD score is associated with
osteopenia and advanced osteoporosis, BMD screening a T-score equal to or above 1.0. In contrast, “low bone
should be repeated every 4 and 2 years, respectively.19 mass” or osteopenia corresponds to a T-score between 1
State of the Art Imaging of Osteoporosis 417

and 2.5 while a diagnosis of osteoporosis corresponds to a

T score less than or equal to 2.5. The category of severe/
established osteoporosis is assigned to those with a T-score
less than or equal to 2.5 in the presence of one or more fra-
gility fractures.25 For other patient populations, namely pre-
menopausal women, men of age less than 50 years, and
children, a Z-score is instead reported to compare a patient’s
BMD to that of an age- and gender-matched population. The
diagnostic categories for osteoporosis as defined by the
WHO are not applied to these specific populations, however
the International Society for Clinical Densitometry (ISCD)
categorizes a Z-score of 2.0 or lower as “below the expected
range for age” and a Z-score above 2.0 as “within the
expected range for age” within their recommendations.26
Advantages of DXA include low radiation requirements,
low cost, quick image acquisition, and high precision. DXA
scans help evaluate fracture risk and are a component of the
Fracture Risk Assessment Tool, which integrates BMD values
with known clinical risk factors to calculate a 10-year proba-
bility of major osteoporotic or hip fracture.27 Follow-up
DXA scans are routinely used in monitoring BMD changes in
response to treatment, usually after 1 to 2 years.28 Using
BMD as a surrogate outcome for fracture reduction showed a
strong association between treatment-related changes in
BMD and reductions in vertebral, hip, and non-vertebral
fractures a meta-analysis of 23 randomized placebo-con-
trolled fracture outcome.29
It is of utmost importance that the image obtained via a
DXA scan is of high quality and done with accuracy and pre-
cision so that any changes detected are attributable to biolog-
ical factors and not to testing errors.30 Among the most Figure 1 Osteoporotic compression fracture of T12 in a 92 years-old
female presenting with back pain. Note that the average L1 anterior
important factors for precision assessment are patient posi-
vertebral body Hounsfield unit in this case (measured at 280 HU) is
tioning and correct scan analysis and ROI definition which
not a reliable indicator of bone density due to sclerotic changes of
requires skilled radio-technologists and radiologists.30,31 The the endplates. This same process potentially may result in false neg-
International Society for Clinical Densitometry (ISCD) rec- ative results in DXA, which emphasizes the need to review the
ommends that patients have their BMD measurements done images obtained for use in DXA before interpretation of semiquanti-
on the same instrument to avoid differences in interpretation tative data.
due to manufacturing/model variabilities.32
The key limitations of DXA stem from its inherent nature increasing attention being paid towards CT and MRI techni-
as a two-dimensional methodology. BMD measurements ques due to their ability to evaluate the 3-dimensional struc-
taken from the lumbar spine are sensitive to morphologic ture of bone.
abnormalities such as prior compression fractures, vertebro-
plasties, laminectomies, or degenerative changes that may
result in artificial changes in BMD (Fig. 1). Adequate quality
control should then be performed as to exclude such altered
Computed Tomography
or deformed vertebral bodies from the BMD measurement, Osteoporosis can also be evaluated using computed tomogra-
although a minimum of two vertebral bodies must be phy (CT). CT is particularly ideal in evaluating bone integrity
included within the analysis.33 It is also important that the due to the imaging modality’s high spatial resolution.9 Imag-
presence of artifacts, such as overlying calcifications, surgical ing findings such as low attenuation and loss of cortical and
material, or retained contrast medium, is acknowledged and trabecular volume can be suggestive of osteoporosis.9
reported alongside any excluded site with the reasoning for CT is particularly advantageous as an imaging modality
the decision.32 Furthermore, the BMD measurements pro- due to its availability, ease, and frequency of use in evaluating
vided by DXA do not provide information regarding changes pathologies.34 Moreover, evaluation by CT avoids the effect
in cortical or trabecular bone structure nor do they allow for of confounding factors such as body habitus (although CT
the distinction of changes in BMD attributable due to bone may be affected by this to a certain extent), calcifications,
geometry or a true increase in bone density.17 With the and degenerative changes of the spine, which often pose
developments in understanding of how central a role micro- problems for DEXA.35 Opportunistic CT may be particularly
architecture plays in determining bone strength, there is useful as an alternate imaging modality when DXA cannot be
418 M. Chen et al.

used, as in, for example, patients with scoliosis or calcifica-

tions from chronic disease, and degenerative changes and
osteophytes, which can artificially elevate DXA results.36
Multiple studies have shown that trabecular attenuation of
the lumbar vertebrae, in particular L1, can be used as a surro-
gate for bone marrow density, especially since the trabecular
pattern in the lumbar vertebrae is relatively uniform.34 While
other studies have assessed the possibility of using other
bones such as the frontal bone, proximal humerus, distal
radius, femoral head and neck, distal tibia, and talus, there
are disadvantages to using non-axial bones. These include
decreased frequency of scans of the appendicular skeleton,
decreased frequency of osteoporotic fractures in the appen-
dicular skeleton, and increased variability of the trabecular
bone HU in other bones of the body.34
Figure 2 On the left is a CT of an 81 years-old female with an aver-
While prior studies have shown no significant difference in age L1 anterior vertebral body density of 81 HU, indicating osteope-
CT attenuation of lumbar vertebra, the L1 vertebra has been nia. This is an example of opportunistic screening for osteoporosis
traditionally used since measures taken from L1 are as accu- by abdominal CT. Compare this image to that on the right, which
rate as measures obtained at other vertebral levels, and L1 is shows the average L1 anterior vertebral body Hounsfield unit in a
easily identifiable, resulting in improved reproducibility of 33 years-old male, who presented for evaluation of a sport injury.
measurements.37 Moreover, L1 is often not affected by The average value is 170 HU, indicating no osteopenia.
degenerative changes seen in lower lumbar levels.35 L1 is
also included on many general CT examinations such as of
the thoracolumbar spine, chest, and abdomen. If L1 is unable Using CT attenuation to infer bone quality could triage
to be used (due to compression fracture, lesion such as enos- patients at risk for osteoporosis and fragility fracture by capi-
tosis or hemangioma, or artifact), then an adjacent level can talizing on preexisting data from the 80 million CT scans per-
be measured.34 formed annually in the U.S.41 Patients who may not qualify
Jang et al.38 showed mean L1 attenuation decreases line- for outpatient DXA may have prior CT scanning due to prior
arly at 2.5 HU per year, and this study also established nor- injuries or pathologies (Fig. 2).36 Using diagnostic CT scans
mative ranges of L1 vertebra trabecular attenuation across to directly measure and compare HU values can potentially
different ages at 120kV. Given their findings, Jang et al.38 be used for opportunistic osteoporosis screening, but it is
used a cutoff of L1 values less than 100 HU as concerning for not currently ready for clinical implementation. HU values
osteoporosis and higher fracture risk. Pickhardt et al.39 vary extensively among different machines, and therefore
showed that in terms of distinguishing osteoporosis from this diagnostic entity is not broadly applicable given its poor
osteopenia on CT, an L1 CT-attenuation less than or equal to replicability.36 Moreover, kilovolt settings have a large effect
160 HU was 90% sensitive, a threshold of 110 HU was on attenuation values, with mean attenuation values decreas-
greater than 90% specific. Their study established an optimal ing as the tube voltage increases. Garner et al.42 showed a sig-
L1 CT-attenuation threshold of 135 HU. Graffy et al.40 in nificant relative decrease in L1 trabecular attenuation at
their study correlating L1 trabecular attenuation with preva- 100 kV versus 140 kV by 45.5 HU.
lence of vertebral body fractures in senior population, sug-
gest a threshold of 90 HU for determining risk of
osteoporotic vertebral fractures, which showed sensitivity,
specificity and a corresponding AUROC of 86.9%, 83.9%,
0.895, respectively. In a more recent study, Vadera et al.37 Quantitative CT and
showed that in a group of approximately 500 patients with
prior DEXA scans, L1 CT-attenuation values were signifi-
Multidetector CT
cantly different between DEXA-defined groups of osteoporo- Quantitative CT (QCT), first developed in the 1970s,
sis, osteopenia, and normal bone density. Overall, these requires the use of phantom calibration and standardized
studies showed, somewhat intuitively, that lower HU thresh- software to calculate either volumetric BMD (in g/cm3) or
old has higher specificity and lower sensitivity for diagnosing areal BMD (in g/cm2).43 Correctly performed QCT examina-
osteoporosis, while a higher HU threshold has higher sensi- tion requires relatively low radiation exposure (»50 mSv)
tivity and lower specificity. compared with conventional spinal radiographs or standard
Studies have shown that administration of IV contrast did CT studies.44
not affect interpretation of L1 trabecular attenuation.38,39 QCT can be performed as follows: a standard CT scanner
Anterior wedging, retropulsion of the posterosuperior verte- can be used for QCT through production of volumetric
bral body cortex, and vacuum phenomenon in the spine are measurements through low-dose scans of either the hip or
common findings in CT associated with osteoporotic spine. Another less common way to perform this imaging is
fractures.9 through use of an extremity scanner, which will allow for
State of the Art Imaging of Osteoporosis 419

obtaining of QCT through high-resolution peripheral quanti- may be at higher risk for fragility fractures and may therefore
tative computed tomography, also known as (HR-)pQCT.9 benefit from increased monitoring of therapy. A drawback of
The femoral neck qCT BMD can be used to determine frac- use of multidetector CT is the large amount of radiation expo-
ture risk using the FRAX tool, with the use of 120 and 80 sure to achieve adequate spatial resolution and image quality,
mg/cm3 as a cutoff. A value <80mg/cm3 corresponds to which limits clinical use of multidetector in osteoporosis at this
osteoporosis and a value between 80 and 120 mg/cm3 corre- time.54,55
sponds to osteopenia.41,43 Finite element analysis (FEA) measures
the stress and strain experienced by a bone undergoing an exter-
nal force based on values calculated from QCT. FEA can thus
predict fracture risk in postmenopausal women and older men,
MR
which correlates to osteoporotic risk factor and is comparable to The use of magnetic resonance imaging techniques in assess-
DXA.9 ing fracture risk, bone microarchitecture, and bone quality is
FEA values can be considered along with other clinical var- attractive for both clinical and research purposes due to the
iables when determining next steps in treatment as well as lack of ionizing radiation associated with this modality.
monitoring improvement and response.41,45 Quantitative CT In high-resolution MR imaging of trabecular bone, the
BMD of the spine can allow for better evaluation of fragility bone matrix is visualized as a signal void within the strong
fractures, as shown in several cross-sectional studies.46,47 surrounding signal of the fatty bone, due to the short T2
Recommendations for the employment of quantitative relaxation time of bone. Early high-resolution MR imaging of
CT instead of DXA are informed by the limitations of trabecular bone was initially performed at peripheral sites
DXA, and are thus in individuals whom are very small or such as the calcaneus and distal radius, although use of this
very large, in older individuals with known advanced imaging modality has subsequently progressed to include the
osteoarthritic or degenerative disease, or in patients where evaluation of more proximal skeletal sites such as the hip
higher sensitivity is needed in monitoring metabolic osse- and proximal femur.50,56,57
ous change, such as in those under treatment with hor- MRI remains the only imaging method able to evaluate the
monal therapeutics.48 bone microarchitecture of the hip in vivo. Given the high sig-
Only 3mSv is necessary to obtain an (HR-)pQCT scan, nal-to-noise constraints of trabecular spine imaging, the abil-
which can be used to acquire volumetric BMD and bone ity to obtain high-resolution images has been limited.58
strength through the generation of both 3D cortical and tra- However, measurements derived from MR-based trabecular
becular bone compartment images.9 Trabecular bone stan- structure analyses have been shown to correlate with in vitro
dard analysis allows for physicians to evaluate the average histology, micro-CT, and biomechanical strength. Longitudi-
number of trabeculae, trabecular thickness, and trabecular nal studies have demonstrated that using MRI to monitor the
separation. Cortical bone analysis measuring cortical thick- effect of therapeutic interventions is feasible.48 FEA has also
ness and cortical porosity can be used to further evaluate been applied to MRI data for the assessment of bone mechan-
patients who may otherwise have elevated BMD on DXA, ical competence and the measurement of the mechanical
such as those with diabetes.9,49 HR-pQCT, in comparison properties of bone including stiffness and strength.59
with multidetector CT and MR imaging, has higher signal-to- Other advanced MRI techniques such as quantitative sus-
noise ratio and spatial resolution. However, its use is limited ceptibility mapping (QSM), magnetic resonance spectros-
to peripheral skeletal sites, which limits this method in its copy, and chemical shift encoding-based water-fat MRI
ability to analyze bone quality in the lumbar spine or proxi- (CSE-MRI) enable improved quantitative assessment of non-
mal femur, which are both common sites for osteoporotic mineralized bone compartments.58 CSE-MRI imaging utilizes
fragility fractures.50 Areas which can be analyzed in a singular the known difference in resonance frequencies between water
pass of HR-pQCT are restricted to portions of the distal and fat to allow for the separation of water and fat signal into
radius and tibia. Furthermore, the scanner tube utilized in individual images. This method allows for spatially resolved
HR-pQCT has a limited life span. Finally, the motion artifacts quantification of the bone marrow fat fraction (BMFF).60
that can mar HR-pQCT scans can end up limiting morpho- Indirect MRI methods, such as proton-magnetic resonance
logic analysis of bone architecture.48 spectroscopy (1H-MRS), point-resolved spectroscopy
Use of multidetector CT is standard in clinical practice, (PRESS), and stimulated echo acquisition mode (STEAM)
especially given its superior spatial resolution when compared can also be used in evaluating bone quality and structure,
with prior iterations of spiral CT scanners; however, the spatial through such special properties as quantification of bone
resolution of multidetector CT is still limited when utilized to marrow fat and ability to characterize central skeleton bone
image trabecular bone given a minimum section thickness marrow fat spectrum.60,61 MR spectroscopy can quantify the
requirement of 0.6 mm, at which individual trabeculae may levels of water and fat present in bone marrow, allowing
still be subject to partial volume effects.51 Trabecular bone additional extraction of information on lipid composition,
parameters obtained with this technique have still been shown such as unsaturation and polyunsaturation levels.62 Diffu-
to correlate with those determined on contact radiographs sion-weighted imaging and dynamic contrast-enhanced MRI
from histologic bone sections and micro-CT.52,53 have also shown quantitative differences in evaluating
Multidetector CT has access to the spine and proximal femur, between acute osteoporotic vertebral fractures and normal
an area where HR-pQCT scanning is limited, but where patients appearing vertebrae.63
420 M. Chen et al.

More recent studies have focused on evaluation of the A meta-analysis by Moayyeri et al.69 demonstrated an asso-
non-mineralized component of bone, such as bone marrow ciation between risk of hip fractures and QUS parameters
adipose tissue (BMAT), in helping determine bone health, taken at the calcaneus, independent of the risk estimates pre-
where increased marrow fat is associated with lower bone dicted from hip DXA measures. Findings in additional stud-
density and vertebral fracture. Advances toward the quantita- ies, the Canadian Multicenter Osteoporosis Study, have
tive MRI assessment of bone marrow composition raises the described similar correlation in fracture risk and QUS in
possibility of utilizing marrow fat as a novel treatment target three additional skeletal sites (distal radius, tibia, and pha-
for osteoporosis therapies.64 T1-weighted MRI has previously lanx).
been used with the application of thresholding methods to The acceptance of quantitative US for use in clinical prac-
quantify BMAT volume and has demonstrated an inverse tice has been limited due to the wide diversity and heteroge-
relationship between BMAT and BMD measured by DXA in neity of quantitative US devices, variations in measuring and
the hip and lumbar spine, which exists independently of age, reporting bone parameters, and as well as variable measure-
sex, ethnicity, and body composition.65 ment sites.61 Therefore, while quantitative US has been
The current spatial resolution achievable with MRI shown to distinguish between those with and without fragil-
results in substantial partial volume effects, and the long ity fractures as well as to predict fracture risk, it is currently
acquisition times make current imaging susceptible to not an established method of osteoporosis diagnosis or treat-
motion artifacts.48 Another obstacle to the usage of MRI ment monitoring according to the WHO classification.48
to evaluate bone structure is due to the signal void corti- However in settings where central DXA is inaccessible, QUS
cal bone produces on conventional MRI pulse sequences has been recognized by the ISCD as having utility, in con-
due to its very short T2 relaxation time. This has led to junction with clinical risk factors, in identifying subjects who
the development of novel sequences with submillisecond have a sufficiently high probability of osteoporotic fracture
echo times, such as ultrashort echo time (UTE) or zero and should initiate treatment using device-specific
echo time (ZTE) allowing for the direct MRI evaluation of thresholds.70
cortical and trabecular bone.61,66 These techniques have
driven advances in the field of quantitative MRI imaging
in their ability to estimate different components of bone
and predict its mechanical properties.66
Bone Radiotracers
After the introduction of radioisotopes for medical use in the
early 20th century, the first bone-targeting agents consisted
mainly of strontium and calcium isotopes.71,72 In 1942,
Ultrasound Treadwell et al. introduced the use of 45Ca and 89Sr, both
Quantitative ultrasound is a low-cost technique free of ioniz- pure beta-emitters, to characterize abnormal calcium metab-
ing radiation that has value in predicting fracture risk, partic- olism in bone tumors73; however, investigations into 47Ca
ularly in clinical settings and areas of the world with limited and 85Sr found these isotopes more desirable for imaging
access to central DXA. The use of quantitative ultrasound because their gamma emissions are more easily detected by
relies on measurement of the propagation of ultrasound external instruments.74-76 Given the long half-life of 85Sr at
waves through the bone. The velocity of transmission of 65 days resulting in high radiation dose to the patient, start-
these waves as well as signal amplitude are used to character- ing in 1964, there was transient interest in skeletal scintigra-
ize these measurements. Thus the most commonly reported phy with 87mSr, which has a half-life of 2.8 hours.77
parameters in QUS are speed of sound and broadband ultra- Although 87mSr was conveniently produced with a generator
sound attenuation.48,67 Speed of sound (SOS, in units of rather than a cyclotron, a major limitation in its clinical appli-
meter/seconds) is defined by the transmission time of the cation for bone imaging was persistent blood pool activity,
sound waves divided by the length of the body part studied. causing false-positive findings in highly vascular soft tissue
Broadband attenuation of sound (BUA) refers to the slope lesions and false-negative findings when low bone-to-back-
between sound wave attenuation and frequency as the waves ground uptake obscured osseous pathology.78
pass through measured bone. Composite parameters derived The advent of 99mTc-labeled bone agents in 1971, start-
from BUA and SOS, such as Stiffness Index (SI) or Quantita- ing with 99mTc-polyphosphate,79 represented such a major
tive Ultrasound Index (QUI) may also be reported.68 improvement in image quality that 87mSr quickly became
QUS can be performed at a number of different skeletal obsolete.72 Soon after in 1975, bone scintigraphy (BS) with
sites, such as the calcaneus, phalanges, and tibia. However, 99mTc-methylene diphosphonate (MDP) was first described,
clinical use of QUS in osteoporosis management is currently demonstrating higher osseous uptake and faster blood clear-
only validated in the heel.26 The pulsed waves produced by ance than radiolabeled pyrophosphate and polyphosphate,80
QUS devices are between the frequencies of 500 kHz and helping to solidify its position as the dominant bone agent in
1.25 MHz, which is much lower than the usual frequencies nuclear medicine today. Reflecting skeletal sites with active
used in imaging US.48 QUS findings in osteoporotic bone mineralization, 99mTc-MDP is thought to bind to hydroxyap-
demonstrate a decreased speed of sound as well as increased atite by chemisorption81,82; however, binding of the tracer to
broadband ultrasound attenuation, and has demonstrated plasma proteins ensures that a portion of the tracer will
strong power in predicting osteoporotic fractures.48 remain in the blood pool rather than depositing on
State of the Art Imaging of Osteoporosis 421

mineralizing bone.83 Subsequent decay of 99mTc, which

occurs with a half-life of 6 hours and detected as a 140 keV
photopeak by the gamma camera, allows for the creation of
images demonstrating tracer distribution in vivo.84 In addi-
tion to planar scintigraphy, 99mTc-labeled agents are also
amenable to imaging with single-photon emission computed
tomography (SPECT), which produces cross sectional images
by reconstructing two-dimensional images acquired from
multiple angles.85 Because imaging with 99mTc relies on the
gamma ray produced directly by its radioactive decay, posi-
tion emission tomography (PET) imaging, in which the scan-
ner detects pairs of photons produced by electronpositron
annihilations, is impossible with 99mTc-labeled agents. The
current limits of technology allow a much higher spatial reso-
lution for PET, in which activity is localized based on coinci- Figure 3 Illustration demonstrating the relative time course of changes
dence events,86 compared to SPECT, in which collimation evident by structural (blue line) and molecular (red line) imaging
causes significantly lower detection efficiency.87 Once proto- techniques. While metabolic changes initiate the disease process,
types for the first integrated SPECT/CT and PET/CT scanners examining structure with modalities such as DEXA and CT may not
were built in the 1990s, hybrid imaging with both SPECT/ be sensitive enough to show any abnormality until months or years
CT and PET/CT entered routine clinical use in the following later, after significant bone loss has occurred. Molecular imaging offers
decade.88,89 a noninvasive approach to examine the ongoing biological processes
Although classified as a PET radiotracer, the position-emit- that drive these macroscopic changes, which may allow timely inter-
ting 18F-sodium fluoride (NaF) was first used for skeletal vention and monitoring of treatment in a patient who would other-
wise develop significant risk of fracture. Image reprinted without
scintigraphy in 1962,90 before the invention of PET resulting
changes from Park et al.53 under the Creative Commons Attribution
from continuous improvements in positron-based imaging
4.0 International License (CC BY). https://creativecommons.org/
techniques throughout the 1970s and 1980s.91,92 Even after licenses/by/4.0/. (Color version of figure is available online.)
advances in imaging technology had successfully achieved
state-of-the-art metabolic bone imaging with NaF-PET/CT,
99m
Tc-MDP BS maintained clinical predominance on the NaF production as an intermediate, NaF-PET/CT has the
basis of lower cost and higher convenience associated with opportunity to match its growing interest among nuclear
the generator-produced 99mTc compared to the cyclotron- medicine physicians by achieving wider clinical use in the
produced 18F. High osseous uptake of the dissociated 18F- future.
ion, which converts hydroxyapatite to fluorapatite via
exchange with the hydroxyl group, combined with rapid
plasma clearance allow for superior image quality less than
1 hour after tracer administration.93 Importantly, protein
Radiotracer Quantification
binding is not encountered when imaging with NaF. 18F, Osteoporosis, as a systemic disorder, is expected to cause dif-
with a half-life of 109.8 minutes, decays primarily by posi- fuse abnormal tracer activity best determined by quantitative
tron emission and produces stable 18O. The PET scintillation analysis rather than focal findings that can be easily discerned
detectors, organized in a ring, record the 511 keV photons visually. Thus, quantification will play a key role in osteopo-
produced by electronpositron annihilation and localize the rosis evaluation by molecular imaging.96-98 An early example
tracer along the line-of-response, assuming two signals of such a technique is the whole-body retention test,
detected within nanoseconds of one another are true coinci- described in 1978 by Fogelman et al.,99 who used a whole-
dences.94 Mean positron range for 18F, defined as the dis- body monitor to measure 99mTc-hydroxyethylidene
tance traveled by the positron prior to annihilation, is diphosphonate activity up to 24 hours after tracer adminis-
estimated at approximately 0.6 mm in water, which is lower tration.99 Although no significant difference in whole-body
than other radionuclides used in PET imaging such as 11C retention was noted between normal subjects and osteoporo-
(1.0 mm), 68Ga (2.7 mm), and 82Rb (5.3 mm).95 Positron sis patients, patients with renal osteodystrophy, Paget’s dis-
range creates a physical limit to the maximum spatial resolu- ease, osteomalacia, and primary hyperparathyroidism all
tion theoretically achievable by PET; despite this fact, clinical demonstrated increased retention compared to controls. Sub-
PET systems outperform clinical SPECT systems in terms of sequent attempts to quantify activity of 99mTc-labeled tracers
spatial resolution by a considerable margin. Therefore, excel- using different equipment and techniques have been met
lent clearance to bone in the absence of protein binding, with moderate success,100-102 but lack the reliability, conve-
combined with the advantages of PET imaging over SPECT, nience, and straightforward approaches to quantification
make NaF-PET/CT the prime imaging modality for observing only possible through PET/CT.
bone metabolism in vivo (Fig. 3). As the availability of NaF Calculation of the NaF bone plasma clearance (Ki) has
increases alongside the ever-expanding production and dis- been successfully employed in discriminating baseline bone
tribution of 18F-fluorodeoxyglucose (FDG), which requires metabolism in healthy and osteoporotic patients as well as
422 M. Chen et al.

showing effects of osteoporosis medical interventions.103-105 significant correlation between NaF uptake avidity and histo-
Derived from compartmental modelling, Ki was originally morphometric parameters such as bone formation, hinting at
calculated using 60-minute dynamic PET imaging with the possibility of replacing invasive biopsy procedures with
approximately 25 venous samples required during the proce- molecular imaging. 124,125 For the next steps for the incorpo-
dure to use in the arterial input function.106 Attempts to sim- ration of NaF-PET/CT in clinical settings, it would be of cru-
ply the approach by using static rather than dynamic cial value to confirm whether NaF uptake values not only
imaging nonetheless require venous sampling.107-109 Stan- correlate with bone density and osteoporosis status but also
dardized uptake value, which is widely used with other PET predict future fractures, which cause morbidity and
tracers such as FDG, does not require venous sampling nor decreased quality of life in patients with metabolic bone dis-
dynamic imaging. Instead, the measured activity is normal- orders.
ized to injected dose body weight and can be expressed in In the realm of clinical trials and disease management,
the units g/mL. Although SUV has been criticized as being NaF-PET/CT evaluates patients’ response to anti-osteoporosis
less rigorous than Ki,110 studies have shown that SUV is not therapy possibly even earlier than traditional biomarkers.
only more convenient, but similar in precision and accu- One of the first study demonstrating the potential of NaF-
racy.111-113 Therefore, if NaF-PET/CT is to be employed clin- PET in assessing therapy response was one by Frost et al.105
ically to assess osteoporosis, calculation of SUV rather than showing that mean vertebral Ki significantly decreased 6
Ki is the logical choice. months after risedronate in patients with T score of less than
2, highlighting the new potential role of NaF-PET as a non-
invasive biomarker. Since then, numerous studies have dem-
onstrated the role of NaF-PET/CT for detection of response
Clinical Application of NaF-PET/ to treatment, such as the anabolic agent teriparatide, as early
as 12 weeks since starting therapy.126,127 Interestingly, teri-
CT paratide had differential effect depending on the skeletal site
NaF-PET/CT has been used to evaluate for osteoporosis pri- with most increased NaF SUVmean observed in the femoral
marily in the spine, femoral neck, and the hip.114-116 In shaft and least at the spine, highlighting the sensitivity of
terms of the technique, the method involves drawing a region NaF-PET/CT for regional differences in bone turnover.127
of interest on the trabecular space and avoiding the cortical Another unique study by Uchida et al.128 revealed that in 24
bone, as the trabecular bone is most metabolically active.46 postmenopausal women with glucocorticoid-induced osteo-
Furthermore, involvement of the whole bone in analysis can porosis, NaF uptake significantly decreased in the lumbar
lead to falsely elevated uptake value by the inclusion degen- spine after 3 months of treatment with alendronate while no
erative changes demonstrating focal avidity.117 Similar meth- significant change was seen in bone mineral density or serum
ods have been established to assess for bone mineral density bone specific alkaline phosphate (Fig. 4), highlighting the
using CT that is simple to perform.118 SPECT and PET also sensitivity of NaF-PET/CT for biochemical changes occurring
show promise in allowing physicians to identify populations in early therapy. As most studies involving NaF-PET and
at risk of bone loss and determine when to start appropriate treatment response have used bisphosphonates,
prophylaxis through the quantification of burn turnover.119
Various studies have confirmed that NaF uptake at skeletal
sites most commonly affected by osteoporosis, such as the
spine, correlate with bone mineral density and decrease with
aging.120-122 In a study of 88 healthy volunteers with age
spectrum ranging from 20 to 70 years, NaF uptake as mea-
sured by SUVmean at the L1-L4 trabecular spine decreased
with age and correlated with bone density as measured by
Hounsfield Unit value.115 Rhodes et al.52 derived a variable
called “bone metabolism score” from the NaF uptake meas-
urements in the femoral neck that correlated with age and T-
scores. A retrospective analysis of 63 lumbar spines aggre-
gated from various trials revealed that NaF uptake correlates
with bone mineral density as measured by DEXA scans.114
Lastly in a study involving 72 postmenopausal women, NaF
activity as measured by Ki at the lumbar spine in osteopo-
Figure 4 Sagittal (A) and coronal (B) NaF-PET images showing
rotic women was significantly decreased compared to those
regions of interest used to calculate SUV in the L1L5 vertebral
of normal or osteopenic group.104
bodies and left femoral neck, respectively.58 This research was origi-
The mechanism for decreased NaF uptake with osteoporo- nally published in JNM. Uchida K, Nakajima H, Miyazaki T, Yayama
sis likely involves decreased bone perfusion as well as T, Kawahara H, Kobayashi S, Tsuchida T, Okazawa H, Fujibayashi
decreased osteoblastic activity with aging.123 Studies of dialy- Y, Baba H. Effects of alendronate on bone metabolism in glucocorti-
sis patients who underwent NaF-PET/CT scans and bone coid-induced osteoporosis measured by 18F-fluoride PET: a pro-
biopsy for the evaluation of a bone mineral disorder found a spective study. J Nucl Med. 2009 Nov;50(11):1808-14. Ó SNMMI.
State of the Art Imaging of Osteoporosis 423

investigations involving novel anti-osteoporosis therapy such 8. Rosen CJ, et al: The epidemiology and pathogenesis of osteoporosis.
as denosumab would be of interesting future direction.129,130 In: Feingold KR, Anawalt B, Blackman MR, et al (eds): Endotext [Inter-
net], South Dartmouth (MA): MDText.com, Inc., 2000 [Accessed
There remains further research to be performed for the
October 8, 2023]. Available from: http://www.ncbi.nlm.nih.gov/
adoption of NaF-PET/CT in assessing for osteoporosis. As books/NBK279134/
NaF-PET/CT have higher radiation exposure compared to 9. Garcia D, Roy SG, Hayeri R: Osteoporosis: Diagnostic imaging and
traditional imaging modalities used for fracture risk calcula- value of multimodality approach in differentiating benign versus path-
tion such as the DXA scan, most preliminary use of NaF- ologic compression fractures. In: Gholamrezanezhad A, Assadi M,
PET/CT will be one that is opportunistic in patients who Jadvar H (eds): Radiology-Nuclear Medicine Diagnostic Imaging, 1st
ed., Wiley, 659-670, 2023. [Accessed October 7, 2023]Available from:
receive the PET scan for oncological purposes.131 For https://onlinelibrary.wiley.com/doi/10.1002/9781119603627.ch23
instance, there is lack of research investigating the potential 10. Zou X, Ma Z, Liu X, et al: Risk factors and prognostic value of osteopo-
of NaF-PET in specific high-risk disease populations such as rosis in hospitalized patients with bronchiectasis. BMC Pulm Med 23
multiple myeloma, where the use of NaF-PET for disease (1):55, 2023
assessment is currently being investigated.132-135 Other dis- 11. Kling JM, Clarke BL, Sandhu NP: Osteoporosis prevention, screening,
and treatment: A review. J Women’s Health 23(7):563-572, 2014
eases with increased risk for osteoporosis where opportunis- 12. Raisz LG: Screening for osteoporosis. N Engl J Med 353(2):164-171,
tic screening with NaF-PET/CT may be of benefit include 2005
hyperparathyroidism, chronic kidney disease, ankylosing 13. Daly RM, Rosengren BE, Alwis G, et al: Gender specific age-related
spondylosis, and rheumatoid arthritis.136-139 changes in bone density, muscle strength and functional performance
in the elderly: A-10 year prospective population-based study. BMC
Geriatr 13(1):71, 2013
14. US Preventive Services Task Force, Curry SJ, Krist AH, et al: Screening
for osteoporosis to prevent fractures: US Preventive Services Task
Summary Force Recommendation Statement. JAMA 319(24):2521, 2018
Given the increasing prevalence of osteoporosis in aging pop- 15. Bernabei R, Martone AM, Ortolani E, et al: Screening, diagnosis and
treatment of osteoporosis: A brief review. Clin Cases Miner Bone
ulations worldwide, it is paramount that continued research
Metab 11(3):201-207, 2014
be performed into emerging methods of evaluation and 16. Oliveira MAD, Moraes R, Castanha EB, et al: Osteoporosis screening:
screening. DXA remains the gold-standard in osteoporosis Applied methods and technological trends. Med Eng Physics
diagnosis and management due to its low cost, imaging 108:103887, 2022
speed, and accessibility, but within CT, MRI, US and in par- 17. Kazakia GJ, Majumdar S: New imaging technologies in the diagnosis of
osteoporosis. Rev Endocr Metab Disord 7(12):67-74, 2006
ticular, nuclear medicine, recent advances continue to prove 18. Viswanathan M, Reddy S, Berkman N, et al. Screening to prevent osteopo-
promising for future adaptation into clinical practice. While rotic fractures: An evidence review for the U.S. preventive services task
not currently utilized in clinical and primary-care settings, force [Internet]. Rockville (MD): Agency for Healthcare Research and
use of NaF-PET and NaF-PET/CT in trials and in opportunis- Quality (US); 2018 [Accessed October 8, 2023]. (U.S. Preventive Services
tic settings show high sensitivity and favorable indicators of Task Force Evidence Syntheses, formerly Systematic Evidence Reviews).
Available from: http://www.ncbi.nlm.nih.gov/books/NBK532075/
use in early detection. 19. Screening and monitoring tests for osteopenia /osteoporosis j Wash-
ington State Health Care Authority [Internet]. [Accessed October 8,
2023]. Available from: https://www.hca.wa.gov/about-hca/programs-
and-initiatives/health-technology-assessment/screening-and-monitor-
Declaration of Competing ing-tests-osteopenia/osteoporosis.
Interest 20. Guermazi A, Mohr A, Grigorian M, et al: Identification of vertebral
fractures in osteoporosis. Semin Musculoskelet Radiol 06(3):241-252,
The authors declare that they have no known competing 2002
21. Dimai HP, Ljuhar R, Ljuhar D, et al: Assessing the effects of long-term
financial interests or personal relationships that could have
osteoporosis treatment by using conventional spine radiographs:
appeared to influence the work reported in this paper. Results from a pilot study in a sub-cohort of a large randomized con-
trolled trial. Skeletal Radiol 48(7):1023-1032, 2019
22. El Maghraoui A, Roux C: DXA scanning in clinical practice. QJM 101
References (8):605-617, 2008
23. Anwar F, Iftekhar H, Taher T, et al: Dual energy x-ray absorptiometry
1. Johnell O, Kanis JA: An estimate of the worldwide prevalence and dis-
scanning and bone health: The pressing need to raise awareness
ability associated with osteoporotic fractures. Osteoporos Int 17
amongst Pakistani women. Cureus [Internet] 2019. [Accessed October
(12):1726-1733, 2006
7, 2023]; Available from: https://www.cureus.com/articles/22647-
2. Johnell O, Kanis J: Epidemiology of osteoporotic fractures. Osteoporos
dual-energy-x-ray-absorptiometry-scanning-and-bone-health-the-
Int 16(S02):S3-S7, 2005
pressing-need-to-raise-awareness-amongst-pakistani-women
3. Dobbs MB, Buckwalter J, Saltzman C: Osteoporosis: The increasing
24. Haseltine KN, Chukir T, Smith PJ, et al: Bone mineral density: Clinical
role of the orthopaedist. Iowa Orthop J 19:43-52, 1999
relevance and quantitative assessment. J Nucl Med 62(4):446-454,
4. Legroux I, Cortet B: Factors influencing bone loss in anorexia nervosa:
2021
Assessment and therapeutic options. RMD Open 5(2):e001009, 2019
25. Assessment of Osteoporosis at the Primary Health Care Level. WHO
5. Lane NE: Epidemiology, etiology, and diagnosis of osteoporosis. Am J
Collaborating Centre for Metabolic Bone Diseases, University of Shef-
Obstetrics Gynecol 194(2):S3-11, 2006
field Medical School; 2008. 288 p.
6. Hellekson KL: NIH releases statement on osteoporosis prevention,
26. Official Positions [Internet]. ISCD. [Accessed October 8, 2023]. Avail-
diagnosis, and therapy. Am Fam Physician 66(1):161-162, 2002
able from: https://iscd.org/learn/official-positions/
7. Izadyar S, Fazeli M, Izadyar M, et al: Bone mineral density in adult
27. Fracture Risk Assessment Tool (FRAX) [Internet]. [Accessed October
patients with major thalassaemia: Our experience and a brief review of
8, 2023]. Available from: https://frax.shef.ac.uk/FRAX/
the literature. Endokrynol Pol 63(4):264-269, 2012
424 M. Chen et al.

28. Blake GM, Fogelman I: Role of dual-energy x-ray absorptiometry in the 49. Nishiyama KK, Shane E: Clinical imaging of bone microarchitecture
diagnosis and treatment of osteoporosis. J Clin Densitometry 10 with HR-pQCT. Curr Osteoporos Rep 11(2):147-155, 2013
(1):102-110, 2007 50. Krug R, Burghardt AJ, Majumdar S, et al: High-resolution imaging
29. Black DM, Bauer DC, Vittinghoff E, et al: Treatment-related changes in techniques for the assessment of osteoporosis. Radiol Clin N Am 48
bone mineral density as a surrogate biomarker for fracture risk reduc- (3):601-621, 2010
tion: Meta-regression analyses of individual patient data from multiple 51. Link TM, Vieth V, Stehling C, et al: High-resolution MRI vs multislice
randomised controlled trials. Lancet Diabetes Endocrinol 8(8):672- spiral CT: Which technique depicts the trabecular bone structure best?
682, 2020 Eur Radiol 13(4):663-671, 2003
30. Kim HS, Yang SO: Quality control of DXA system and precision test of 52. Issever AS, Vieth V, Lotter A, et al: Local differences in the trabecular
radio-technologists. J Bone Metab 21(1):2, 2014 bone structure of the proximal femur depicted with high-spatial-reso-
31. Garg MK, Kharb S: Dual energy X-ray absorptiometry: Pitfalls in mea- lution MR imaging and multisection CT. Academic Radiology 9
surement and interpretation of bone mineral density. Indian J Endocri- (12):1395-1406, 2002 Dec
nol Metab 17(2):203-210, 2013 53. Diederichs G, Link TM, Kentenich M, et al: Assessment of trabecular
32. Lewiecki EM, Binkley N, Morgan SL, et al: Best practices for dual- bone structure of the calcaneus using multi-detector CT: Correla-
energy x-ray absorptiometry measurement and reporting: International tion with microCT and biomechanical testing. Bone 44(5):976-
Society for Clinical Densitometry Guidance. J Clin Densitometry 19 983, 2009
(2):127-140, 2016 54. Damilakis J, Adams JE, Guglielmi G, et al: Radiation exposure in X-ray-
33. Lorente-Ramos R, Azpeitia-Arm
an J, Mu~ noz-Hern
andez A, et al: Dual- based imaging techniques used in osteoporosis. Eur Radiol 20
energy x-ray absorptiometry in the diagnosis of osteoporosis: A practi- (11):2707-2714, 2010
cal guide. AJR Am J Roentgenol 196(4):897-904, 2011 55. Graeff C, Timm W, Nickelsen TN, et al: Monitoring teriparatide-asso-
34. Boutin RD, Lenchik L: Value-added opportunistic CT: Insights into ciated changes in vertebral microstructure by high-resolution CT in
osteoporosis and sarcopenia. Am J Roentgenol 215(3):582-594, 2020 vivo: Results from the EUROFORS study. J Bone Miner Res 22
35. Smith AD: Screening of bone density at CT: An overlooked opportu- (9):1426-1433, 2007
nity. Radiology 291(2):368-369, 2019 56. Chang G, Boone S, Martel D, et al: MRI assessment of bone structure and
36. Gausden EB, Nwachukwu BU, Schreiber JJ, et al: Opportunistic use of microarchitecture. Magnetic Resonance Imaging 46(2):323-337, 2017
CT imaging for osteoporosis screening and bone density assessment: A 57. Carballido-Gamio J, Folkesson J, Karampinos DC, et al: Generation of
qualitative systematic review. J Bone Joint Surg 99(18):1580-1590, an atlas of the proximal femur and its application to trabecular bone
2017 analysis: Atlas-based trabecular bone analysis. Magn Reson Med 66
37. Vadera S, Osborne T, Shah V, et al: Opportunistic screening for osteo- (4):1181-1191, 2011
porosis by abdominal CT in a British population. Insights Imaging 14 58. Sollmann N, L€offler MT, Kronthaler S, et al: MRI -based quantitative
(1):57, 2023 osteoporosis imaging at the spine and femur. Magnetic Resonance
38. Jang S, Graffy PM, Ziemlewicz TJ, et al: Opportunistic osteoporosis Imaging 54(1):12-35, 2021
screening at routine abdominal and thoracic CT: Normative L1 trabec- 59. Jerban S, Alenezi S, Afsahi AM, et al: MRI-based mechanical compe-
ular attenuation values in more than 20 000 adults. Radiology 291 tence assessment of bone using micro finite element analysis (micro-
(2):360-367, 2019 FEA): Review. Magn Reson Imaging 88:9-19, 2022
39. Pickhardt PJ, Pooler BD, Lauder T, et al: Opportunistic screening for 60. Cordes C, Baum T, Dieckmeyer M, et al: MR-based assessment of bone
osteoporosis using abdominal computed tomography scans obtained marrow fat in osteoporosis, diabetes, and obesity. Front Endocrinol
for other indications. Ann Intern Med 158(8):588, 2013 (Lausanne) Jun 27(7):74, 2016.
40. Graffy PM, Lee SJ, Ziemlewicz TJ, et al: Prevalence of vertebral com- 61. Oei L, Koromani F, Rivadeneira F, et al: Quantitative imaging methods
pression fractures on routine CT scans according to L1 trabecular in osteoporosis. Quant Imaging Med Surg 6(6):680-698, 2016
attenuation: Determining relevant thresholds for opportunistic osteo- 62. Karampinos DC, Ruschke S, Dieckmeyer M, et al: Quantitative MRI
porosis screening. Am J Roentgenol 209(3):491-496, 2017 and spectroscopy of bone marrow. Magnetic Resonance Imaging 47
41. Engelke K, Lang T, Khosla S, et al: Clinical use of quantitative com- (2):332-353, 2018
puted tomography (QCT) of the hip in the management of osteoporo- 63. Biffar A, Sourbron S, Dietrich O, et al: Combined diffusion-weighted
sis in adults: The 2015 ISCD official positions—Part I. J Clin and dynamic contrast-enhanced imaging of patients with acute osteo-
Densitometry 18(3):338-358, 2015 porotic vertebral fractures. Eur J Radiol 76(3):298-303, 2010
42. Garner HW, Paturzo MM, Gaudier G, et al: Variation in attenuation in 64. Li X, Schwartz AV: MRI assessment of bone marrow composition in
L1 trabecular bone at different tube voltages: Caution is warranted osteoporosis. Curr Osteoporos Rep 18(1):57-66, 2020
when screening for osteoporosis with the use of opportunistic CT. Am 65. Shen W, Scherzer R, Gantz M, et al: Relationship between MRI-mea-
J Roentgenol 208(1):165-170, 2017 sured bone marrow adipose tissue and hip and spine bone mineral
43. Zysset P, Qin L, Lang T, et al: Clinical use of quantitative computed density in African-American and Caucasian participants: The CARDIA
tomographybased finite element analysis of the hip and spine in the study. J Clin Endocrinol Metabol 97(4):1337-1346, 2012
management of osteoporosis in adults: The 2015 ISCD official posi- 66. Ma YJ, Jerban S, Jang H, et al: Quantitative ultrashort echo time (UTE)
tions—Part II. J Clin Densitometry 18(3):359-392, 2015 magnetic resonance imaging of bone: An update. Front Endocrinol
44. Guglielmi G, Lang TF: Quantitative computed tomography. Semin (Lausanne) 11:567417, 2020
Musculoskelet Radiol 06(3):219-228, 2002 67. Hans D, Baim S: Quantitative ultrasound (QUS) in the management of
45. Black DM, Greenspan SL, Ensrud KE, et al: The effects of parathyroid osteoporosis and assessment of fracture risk. J Clin Densitometry 20
hormone and alendronate alone or in combination in postmenopausal (3):322-333, 2017
osteoporosis. N Engl J Med 349(13):1207-1215, 2003 68. Chin KY, Ima-Nirwana S: Calcaneal quantitative ultrasound as a deter-
46. Bergot C, Laval-Jeantet AM, Hutchinson K, et al: A comparison of spi- minant of bone health status: What properties of bone does it reflect?
nal quantitative computed tomography with dual energy X-ray absorp- Int J Med Sci 10(12):1778-1783, 2013
tiometry in European women with vertebral and nonvertebral 69. Moayyeri A, Adams JE, Adler RA, et al: Quantitative ultrasound of the
fractures. Calcif Tissue Int 68(2):74-82, 2001 heel and fracture risk assessment: An updated meta-analysis. Osteo-
47. Yu W, Gl€ uer C, Grampp S, et al: Spinal bone mineral assessment in poros Int 23(1):143-153, 2012
postmenopausal women: A comparison between dual X-ray absorpti- 70. Krieg MA, Barkmann R, Gonnelli S, et al: Quantitative ultrasound in
ometry and quantitative computed tomography. Osteoporosis Int 5 the management of osteoporosis: The 2007 ISCD official positions. J
(6):433-439, 1995 Clin Densitometry 11(1):163-187, 2008
48. Link TM: Osteoporosis imaging: State of the art and advanced imaging. 71. DeNardo GL, Volpe JA: Detection of bone lesions with the strontium-
Radiology 263(1):3-17, 2012 85 Scintiscan. J Nucl Med 7(3):219, 1966
State of the Art Imaging of Osteoporosis 425

72. Blake GM, Fogelman I: Bone radionuclide imaging, quantitation and 97. Raynor WY, Borja AJ, Hancin EC, et al: Novel musculoskeletal and
bone densitometry. In: McCready R, Gnanasegaran G, Bomanji JB orthopedic applications of 18F-sodium fluoride PET. PET Clin 16
(eds): A History of Radionuclide Studies in the UK: 50th Anniversary (2):295-311, 2021
of the British Nuclear Medicine Society [Internet], Cham (CH): 98. Zhang V, Koa B, Borja AJ, et al: Diagnosis and monitoring of osteopo-
Springer, 2016 [Accessed 2023 Oct 8]. Available from: http://www. rosis with total-body 18F-sodium fluoride-PET/CT. PET Clin 15
ncbi.nlm.nih.gov/books/NBK500203/ (4):487-496, 2020
73. Treadwell A, Low-Beer BVA, Friedell HL, Lawrence JH: Metabolic 99. Fogelman I, Bessent RG, Turner JG, et al: The use of whole-body reten-
studies on neoplasm of bone with the aid of radioactive strontium. Am tion of Tc-99m diphosphonate in the diagnosis of metabolic bone dis-
J Med Sci 204:521-530, 1942 ease. J Nucl Med 19(3):270-275, 1978
74. Comar CL, Wasserman RH, Ullberg S, et al: Strontium metabolism and 100. Moore AEB, Blake GM, Fogelman I: Quantitative measurements of
strontium-calcium discrimination in man. Proc Soc Exp Biol Med 95 bone remodeling using 99mTc-methylene diphosphonate bone scans
(2):386-391, 1957 and blood sampling. J Nucl Med 49(3):375-382, 2008
75. Bauer GC, Ray RD: Kinetics of strontium metabolism in man. J Bone 101. Shahbazi-Gahrouei D, Damoori M, Tavakoli MB, et al: Comparison of
Joint Surg Am 40-A(1):171-186, 1958 accuracy in calculation of absorbed dose to patients following bone
76. Bauer GC, Wendeberg B: External counting of Ca47 and Sr85 in stud- scan with 99mTc-marked diphosphonates by two different back-
ies of localised skeletal lesions in man. J Bone Joint Surg Br 41-B:558- ground correction methods. J Med Signals Sens 6(1):33-38, 2016
580, 1959 102. Huang K, Feng Y, Liu D, et al: Quantification evaluation of 99mTc-
77. Charkes ND, Sklaroff DM, Bierly J: Detection of metastatic cancer to MDP concentration in the lumbar spine with SPECT/CT: Compare
bone by scintiscanning with strontium 87M. Am J Roentgenol Radium with bone mineral density. Ann Nucl Med 34(2):136-143, 2020
Ther Nucl Med 91:1121-1127, 1964 103. Frost ML, Blake GM, Cook GJR, et al: Differences in regional bone per-
78. Charkes ND: Some differences between bone scans made with /sup fusion and turnover between lumbar spine and distal humerus: (18)F-
87m/Sr AND $sup 85$Sr. J Nucl Med 10:491-494, 1969.. [Accessed fluoride PET study of treatment-naïve and treated postmenopausal
October 8, 2023]; Available from: https://www.osti.gov/biblio/ women. Bone 45(5):942-948, 2009 Nov
4758163 104. Frost ML, Fogelman I, Blake GM, et al: Dissociation between global
79. Subramanian G, McAfee JG: A new complex of 99mTc for skeletal markers of bone formation and direct measurement of spinal bone for-
imaging. Radiology 99(1):192-196, 1971 mation in osteoporosis. J Bone Miner Res 19(11):1797-1804, 2004
80. Subramanian G, McAfee JG, Blair RJ, et al: Technetium-99m-methy- 105. Frost ML, Cook GJR, Blake GM, et al: A prospective study of risedro-
lene diphosphonateA superior agent for skeletal imaging: Compari- nate on regional bone metabolism and blood flow at the lumbar spine
son with other technetium complexes. J Nucl Med 16(8):744-755, measured by 18F-fluoride positron emission tomography. J Bone
1975 Miner Res 18(12):2215-2222, 2003
81. Kanishi D: 99mTc-MDP accumulation mechanisms in bone. Oral Surg 106. Hawkins RA, Choi Y, Huang SC, et al: Evaluation of the skeletal kinetics of
Oral Med Oral Pathol 75(2):239-246, 1993 fluorine-18-fluoride ion with PET. J Nucl Med 33(5):633-642, 1992
82. Toegel S, Hoffmann O, Wadsak W, et al: Uptake of bone-seekers is 107. Blake GM, Puri T, Siddique M, et al: Site specific measurements of
solely associated with mineralisation! A study with 99mTc-MDP, bone formation using [18F] sodium fluoride PET/CT. Quant Imaging
153Sm-EDTMP and 18F-fluoride on osteoblasts. Eur J Nucl Med Mol Med Surg 8(1):47-59, 2018
Imaging 33(4):491-494, 2006 108. Blake GM, Siddique M, Puri T, et al: A semipopulation input function
83. Blake GM, Moore AEB, Park-Holohan SJ, et al: A direct in vivo mea- for quantifying static and dynamic 18F-fluoride PET scans. Nucl Med
surement of 99mTc-methylene diphosphonate protein binding. Nucl Commun 33(8):881-888, 2012
Med Commun 24(7):829-835, 2003 109. Siddique M, Blake GM, Frost ML, et al: Estimation of regional bone
84. Green CH: Technetium-99m production issues in the United King- metabolism from whole-body 18F-fluoride PET static images. Eur J
dom. J Med Phys 37(2):66-71, 2012 Nucl Med Mol Imaging 39(2):337-343, 2012
85. Bartel TB, Kuruva M, Gnanasegaran G, et al: SNMMI procedure stan- 110. Blake GM, Siddique M, Frost ML, et al: Radionuclide studies of bone
dard for bone scintigraphy 4.0. J Nucl Med Technol 46(4):398-404, metabolism: Do bone uptake and bone plasma clearance provide equivalent
2018 measurements of bone turnover? Bone 49(3):537-542, 2011
86. Shukla AK, Kumar U: Positron emission tomography: An overview. J 111. Al-Beyatti Y, Siddique M, Frost ML, et al: Precision of 18F-fluoride PET
Med Phys 31(1):13-21, 2006 skeletal kinetic studies in the assessment of bone metabolism. Osteo-
87. Tsui BM: The AAPM/RSNA physics tutorial for residents. Physics of poros Int 23(10):2535-2541, 2012
SPECT. Radiographics 16(1):173-183, 1996 112. Puri T, Blake GM, Frost ML, et al: Comparison of six quantitative
88. Hutton BF: The origins of SPECT and SPECT/CT. Eur J Nucl Med Mol methods for the measurement of bone turnover at the hip and lumbar
Imaging 41(Suppl 1):S3-16, 2014 spine using 18F-fluoride PET-CT. Nucl Med Commun 33(6):597-
89. Townsend DW: Combined positron emission tomography-computed 606, 2012
tomography: The historical perspective. Semin Ultrasound CT MR 29 113. Frost ML, Siddique M, Blake GM, et al: Regional bone metabolism at
(4):232-235, 2008 the lumbar spine and hip following discontinuation of alendronate
90. Blau M, Nagler W, Bender MA: Fluorine-18: A new isotope for bone and risedronate treatment in postmenopausal women. Osteoporos Int
scanning. J Nucl Med 3:332-334, 1962 23(8):2107-2116, 2012
91. Rich DA: A brief history of positron emission tomography. J Nucl Med 114. Jassel IS, Siddique M, Frost ML, et al: The influence of CT and dual-
Technol 25(1):4-11, 1997 energy X-ray absorptiometry (DXA) bone density on quantitative [18F]
92. Jones T, Townsend D: History and future technical innovation in posi- sodium fluoride PET. Quant Imaging Med Surg 9(2):201-209, 2019
tron emission tomography. J Med Imaging (Bellingham) 4(1):011013, 115. Park PSU, Resto DA, Khurana N, et al: The utility of 18 F-NaF-posi-
2017 tron emission tomography/computed tomography in measuring the
93. Puri T, Frost ML, Cook GJ, et al: [18F] sodium fluoride PET kinetic metabolic activity of the aging spine : Implications for osteoporosis.
parameters in bone imaging. Tomography 7(4):843-854, 2021 Spine (Phila Pa 1976 48(15):1064-1071, 2023
94. Schaart DR: Physics and technology of time-of-flight PET detectors. 116. Ayubcha C, Raynor WY, Borja AJ, et al: Magnetic resonance imaging-based
Phys Med Biol 66(9):09TR01, 2021 partial volume-corrected 18F-sodium fluoride positron emission tomogra-
95. Cal-Gonz
alez J, Herraiz JL, Espa~ na S, et al: Positron range estimations phy in the femoral neck. Nucl Med Commun 42(4):416-420, 2021
with PeneloPET. Phys Med Biol 58(15):5127, 2013 117. Jadvar H, Desai B, Conti PS: Sodium 18F-fluoride PET/CT of bone,
96. Chaudhari AJ, Raynor WY, Gholamrezanezhad A, et al: Total-body joint, and other disorders. Semin Nucl Med 45(1):58-65, 2015
PET imaging of musculoskeletal disorders. PET Clin 16(1):99-117, 118. Schreiber JJ, Anderson PA, Hsu WK: Use of computed tomography for
2021 assessing bone mineral density. Neurosurg Focus 37(1):E4, 2014
426 M. Chen et al.

119. Katal S, Clifford TG, Matcuk G, et al: Clinical applications of PET in 131. Sheppard AJ, Paravastu SS, Wojnowski NM, et al: Emerging role of 18F-
evaluating the aging spine. PET Clin 18(1):39-47, 2023 NaF PET/computed tomographic imaging in osteoporosis: A potential
120. Austin AG, Raynor WY, Reilly CC, et al: Evolving role of MR imaging upgrade to the osteoporosis toolbox. PET Clin 18(1):1-20, 2023
and PET in assessing osteoporosis. PET Clin 14(1):31-41, 2019 132. Sachpekidis C, Hillengass J, Goldschmidt H, et al: Treatment response
121. Reilly CC, Raynor WY, Hong AL, et al: Diagnosis and monitoring of evaluation with 18F-FDG PET/CT and 18F-NaF PET/CT in multiple
osteoporosis with 18F-sodium fluoride PET: An unavoidable path for myeloma patients undergoing high-dose chemotherapy and autolo-
the foreseeable future. Semin Nucl Med 48(6):535-540, 2018 gous stem cell transplantation. Eur J Nucl Med Mol Imaging 44(1):50-
122. Raynor W, Houshmand S, Gholami S, et al: Evolving role of molecular 62, 2017
imaging with (18)F-sodium fluoride PET as a biomarker for calcium 133. Sachpekidis C, Goldschmidt H, Hose D, et al: PET/CT studies of multi-
metabolism. Curr Osteoporos Rep 14(4):115-125, 2016 ple myeloma using (18) F-FDG and (18) F-NaF: Comparison of distri-
123. Park PSU, Raynor WY, Sun Y, et al: 18F-sodium fluoride PET as a bution patterns and tracers’ pharmacokinetics. Eur J Nucl Med Mol
diagnostic modality for metabolic, autoimmune, and osteogenic bone Imaging 41(7):1343-1353, 2014
disorders: Cellular mechanisms and clinical applications. Int J Mol Sci 134. Zadeh MZ, Raynor WY, Seraj SM, et al: Evolving roles of fluorodeoxy-
22(12):6504, 2021 glucose and sodium fluoride in assessment of multiple myeloma
124. Vrist MH, Østergaard AMH, Langaa SS, et al: Bone turnover, minerali- patients: Introducing a novel method of PET quantification to over-
zation, and volume estimated by 18F-sodium fluoride PET/CT and come shortcomings of the existing approaches. PET Clin 14(3):341-
biomarkers in chronic kidney disease: Mineral and bone disorder com- 352, 2019
pared with bone biopsy. Am J Nephrol 53(6):490-502, 2022 135. Zirakchian Zadeh M, Østergaard B, Raynor WY, et al: Comparison of
125. Aaltonen L, Koivuviita N, Sepp€anen M, et al: Correlation between 18F- 18F-sodium fluoride uptake in the whole bone, pelvis, and femoral
sodium fluoride positron emission tomography and bone histomorph- neck of multiple myeloma patients before and after high-dose therapy
ometry in dialysis patients. Bone 134:115267, 2020 May and conventional-dose chemotherapy. Eur J Nucl Med Mol Imaging
126. Frost ML, Moore AE, Siddique M, et al: 18F-fluoride PET as a noninva- 47(12):2846-2855, 2020
sive imaging biomarker for determining treatment efficacy of bone 136. Graf C, Huellner M, Tschopp O, et al: 18F-NaF-PET/CT in patients
active agents at the hip: A prospective, randomized, controlled clinical with primary hyperparathyroidism and brown tumors. J Bone Miner
study. J Bone Miner Res 28(6):1337-1347, 2013 Metab 38(3):299-309, 2020
127. Frost ML, Siddique M, Blake GM, et al: Differential effects of teripara- 137. Vrist MH, Bech JN, Lauridsen TG, et al: Comparison of [18F] NaF
tide on regional bone formation using (18)F-fluoride positron emis- PET/CT dynamic analysis methods and a static analysis method
sion tomography. J Bone Miner Res 26(5):1002-1011, 2011 including derivation of a semi-population input function for site-spe-
128. Uchida K, Nakajima H, Miyazaki T, et al: Effects of alendronate on cific measurements of bone formation in a population with chronic
bone metabolism in glucocorticoid-induced osteoporosis measured by kidney disease-mineral and bone disorder. EJNMMI Res 11(1):117,
18F-fluoride PET: A prospective study. J Nucl Med 50(11):1808- 2021
1814, 2009 138. Son SM, Kim K, Pak K, et al: Evaluation of the diagnostic performance
129. van der Bruggen W, Vriens D, Meier ME, et al: Denosumab reduces of 18F-NaF positron emission tomography/computed tomography in
lesional fluoride skeletal burden on Na[18F]F PET-CT in patients with patients with suspected ankylosing spondylitis according to the Assess-
fibrous dysplasia/McCune-Albright syndrome. J Clin Endocrinol ment of SpondyloArthritis International Society criteria. Spine J 20
Metab 106(8):e2980-e2994, 2021 (9):1471-1479, 2020
130. Fushimi M, Ohbayashi Y, Nakai F, et al: The effect of zoledronic acid 139. Park HJ, Chang SH, Lee JW, et al: Clinical utility of F-18 sodium fluo-
and denosumab on the mandible and other bones: A 18F-NaF-PET ride PET/CT for estimating disease activity in patients with rheumatoid
study. Oral Radiol 38(4):594-600, 2022 arthritis. Quant Imaging Med Surg 11(4):1156-1169, 2021