Dual Energy X-Ray Absorptimetry: Fundamentals, Methodology, and Clinical Applications 411

Home , Bone density

Document downloaded from http://www.elsevier.es, day 22/01/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.

Radiología. 2012;54(5):410---423

www.elsevier.es/rx

UPDATE IN RADIOLOGY

Dual energy X-ray absorptimetry: Fundamentals, methodology,

ଝ

and clinical applications

∗

R.M. Lorente Ramos , J. Azpeitia Armán, N. Arévalo Galeano, A. Munoz˜ Hernández,

J.M. García Gómez, J. Gredilla Molinero

Unidad Central de Radiodiagnóstico de la CAM, Hospital Infanta Leonor, Madrid, Spain

Received 15 March 2011; accepted 27 September 2011

KEYWORDS Abstract Dual-energy X-ray absorptiometry (DXA; DEXA) is the technique of choice to diag-

Dual-energy X-ray nose osteoporosis and to monitor the response to treatment. It is also useful for measuring

absorptiometry; body composition. In recent years, new applications have been developed, including vertebral

Densitometry; morphometry through the study of the lateral spine, prosthesis integration in orthopedics, and

DXA; lipodystrophy in HIV+ patients, although its use in these cases is not well established. DXA

DEXA; densitometry is accurate and precise. It is essential to optimize each step of the diagnostic

Osteoporosis; process, taking care to ensure the best acquisition, image analysis, and interpretation of the

Bone mineral density; results. Thus, to obtain the greatest utility from DXA, radiologists need to know the technique,

Body composition its indications, and its pitfalls. This article reviews the fundamentals, modalities, methods, and

clinical applications of DXA.

PALABRAS CLAVE

Absorciometría con rayos X de doble energía. Fundamentos, metodología

Absorciometría con y aplicaciones clínicas

rayos X de doble

energía; Resumen La absorciometría con rayos X de doble energía (DXA o DEXA) es la técnica de elec-

Densitometría; ción para diagnosticar la osteoporosis y monitorizar la respuesta al tratamiento. Además, es útil

DXA; para estudiar la composición corporal. En los últimos anos˜ han surgido nuevas aplicaciones como

DEXA; la morfometría vertebral, estudiando la columna en visión lateral, la integración de prótesis

Osteoporosis; en ortopedia, o la lipodistroﬁa en los pacientes con infección por VIH, aunque su utilización en

Densidad mineral estos casos no está bien consolidada. En el estudio de la osteoporosis, densitometría es precisa

ósea; y exacta. Para ello, es imprescindible optimizar cada etapa del proceso diagnóstico, cuidando la

Composición corporal adquisición, el análisis de imágenes y la interpretación de los resultados. Por ello, para obtener

la máxima utilidad para el clínico y el paciente, el radiólogo debe conocer la técnica, sus

ଝ

Please cite this article as: Lorente Ramos RM, et al. Absorciometría con rayos X de doble energía. Fundamentos, metodología

y aplicaciones clínicas. Radiología. 2012;54:410---23. ∗

Corresponding author.

E-mail address: [email protected] (R.M.

Lorente Ramos).

Document downloaded from http://www.elsevier.es, day 22/01/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.

Dual energy X-ray absorptimetry: Fundamentals, methodology, and clinical applications 411

indicaciones y las diﬁcultades. El objetivo de este artículo es revisar la DXA, haciendo hincapié

en sus fundamentos, modalidades, metodología y aplicaciones clínicas.

Introduction the analysis algorithm discriminate bone from soft tissue in

a variable way.

The main modalities of DXA in clinical practice are axial

Dual X-ray absorptiometry (DXA), also known as densit-

bone densitometry with stationary scan table, the modality

ometry or dual energy X-ray absorptiometry (DEXA), can

of choice to measure the BMD, and whole body densitometry,

distinguish different body structures. Axial bone densito-

used to assess body composition.

metry of the lumbar spine and hip is the modality most

commonly used in clinical practice. This technique is use-

ful for measuring the bone mineral density (BMD), and from

these data the risk of fracture can be estimated, therapeu-

Bone densitometry with dual energy X-ray

tic decisions can be taken and the response to treatment

1 absorptiometry

can be assessed.

On the other hand, DXA, the least known method for

whole body imaging, allows us to assess the total body Historical perspective

composition. This is particularly useful in patients with

weight disorders secondary to endocrine diseases and in Axial DXA of the lumbar spine and hip is at present the

2 11

pediatric patients with delayed growth. Whole body DXA technique of choice to study osteoporosis, although other

can also be useful to assess lipodystrophy associated with imaging techniques are potentially useful to assess and mea-

3 4 1,7,12

retroviral infections, in the monitoring of arthroplasties sure the bone structure and study its quality (Table 1).

or to determine the cardiovascular risk. Plain radiography is useful to assess bone structure,

DXA is still little known among radiologists, who con- although it cannot measure BMD. Some authors have tried

sider this technique to be more typical of other specialties. to apply digital radiography with dual energy to obtain an

Moreover, DXA tends to be wrongly considered as a rou- estimated measurement of the BMD.

tine and automated technique, unlikely to be optimized Quantitative computed tomography (QCT) of the lum-

and not requiring a radiological report. This is far from the bar spine (central QCT) is performed using conventional

truth. DXA, like any other diagnostic modality, requires an computed tomography (CT) systems. QCT of radius or tibia

adequate indication, careful methodology and precise inter- (peripheral QCT) can be performed using less sophisti-

pretation, which is only possible with appropriate training cated equipment. QCT provides volumetric acquisitions from

and interaction between technicians and radiologists. which BMD can be estimated. Central QCT has advantages

As a consequence, our aim is to examine the current sta- over DXA, since it allows us to differentiate between cortical

tus of DXA, particularly emphasizing its fundamentals, main and trabecular bone, assess the geometry of the vertebrae,

modalities, methodology and clinical applications. and estimate the BMD volumetrically, expressed in g/cm .

The disadvantages of central QCT are the radiation dose and

the lack of validated diagnostic criteria.

High resolution magnetic resonance (MR) imaging may be

Fundamentals and modalities of dual energy X-ray used for assessment of the trabecular structure of periph-

absorptiometry eral bones (calcaneus, distal radius and phalanx). The bone

architecture studied using CT or MR, quantiﬁed in terms

DXA is based on the variable absorption of X-ray by the dif- of scale, shape, anisotropy and connectivity, allows for the

ferent body components and uses high and low energy X-ray assessment of bone strength without considering the BMD.

photons. Depending on the equipment used, these photons Advanced MR techniques such as diffusion, perfusion and

can be obtained using two mechanisms. In some cases, the spectroscopy will most likely provide useful additional infor-

generator emits alternating radiation of high (140 kVp) and mation in the future.

low (70---100 kVp) kilovoltage while moving across the sur- Quantitative ultrasound (QUS) is used for measuring BMD

face of the body to be examined. In others, the generator in the peripheral skeleton, generally at the calcaneus.

emits a constant beam while a rare-earth ﬁlter separates Photonic absorptiometry with iodine-125 (I-125) was ini-

high energy (70 keV) from low energy (40 keV) photons. tially used to study the peripheral skeleton (radius and

The available DXA systems include different types of calcaneus). It was subsequently replaced by dual pho-

hardware (ﬁlters, collimators, detectors) and software tonic absorptiometry that uses gadolinium-153 and may be

(analysis algorithms). The X-ray source can emit a pencil- employed to study the axial skeleton (hip, spine and whole

beam (pinhole collimator), which is registered by a single skeleton).

detector, or a fan beam (slit collimator), which is registered These modalities were later on substituted by X-ray based

by a multiple detector. The latter system reduces the acqui- technology, initially by plain X-ray and subsequently by DXA,

sition time and improves image quality. At the same time, which allowed for the measurement of the axial skeleton.

Document downloaded from http://www.elsevier.es, day 22/01/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.

412 R.M. Lorente Ramos et al.

Table 1 Densitometry modalities.

X-ray Plain X-ray X-ray

QCT Quantitative computed tomography X-ray (CT)

MR Magnetic resonance MR

QUS Quantitative ultrasound Ultrasound

SPA Single photon absorptiometry Radioisotope 125-I

DPA Dual photon absorptiometry Radioisotope Gd-153

SXA Single X-ray absorptiometry X-ray

DXA/DEXA Dual X-ray absorptiometry/Dual energy X-ray absorptiometry X-ray

Equipment Foundation (NOF), osteoporosis affects 10 million Ameri-

cans, but another 34 million are at risk of developing the

Peripheral DXA performed with portable units (such as condition. It is estimated that approximately 50% of women

AccuDXA ) focuses on the study of phalanxes. It is not very over the age of 50 will suffer an osteoporic fracture dur-

1 22

accurate, but its cost is also low. AccuDXA can be used to ing their lives. In Europe, the International Osteoporosis

16 23

select patients likely to be assessed with central DXA on Foundation (IOF) collects data and promotes initiatives in

stationary table or as a substitute where central DXA is not every country. In Spain, approximately two million women

17 24

available. have osteoporosis (26.1% prevalence in women ≥50 years).

Axial DXA of the lumbar spine and femur (central DXA) is In 1994, the WHO introduced the measurement of BMD

the preferred technique to measure BMD because of its good with DXA as the reference standard to quantify osteoporo-

resolution and reliability, rapid acquisition and little radi- sis. Based on a study performed on postmenopausal white

ation. Several devices are commercially available (Lunar, women that revealed a correlation between BMD and risk of

20 −

Hologic, Norland) with different characteristics. This fact fracture, osteoporosis was deﬁned as a ‘‘T-score of 2.5’’.

makes it advisable to perform the follow up of each patient Other values of reference for potentially useful parameters

always in the same unit. The accuracy of bone densitometry were also determined (Table 3). Thus, axial DXA became

performed on a DXA system with stationary table is high, the reference standard for osteoporosis.

with a margin of error of 1---2%. Methodology

Indications

The methodology of bone densitometry with axial DXA

The main use of DXA is the diagnosis of osteoporosis. It requires optimization and careful execution. The impor-

may also predict the risk of fractures, help determine the tance of each step to obtain good results must be

treatment, and monitor the response to treatment. Its highlighted.

indications are set out in the ofﬁcial guidelines of the Inter-

national Society for Clinical Densitometry, which are revised

19,20 Preparation

every two years (Table 2).

In order to properly plan the study, detailed patient infor-

Osteoporosis is the ‘‘reduction in bone mass and increase

mation is required, and both the medical report provided by

of bone fragility which in turn increases the risk of frac-

21 the referring physician as well as the preliminary question-

ture’’. Osteoporosis is a common, often silent disease

naire completed in the diagnostic center. The request form

that can lead to an increased risk of fractures, sometimes

must include the indications for the study, so the areas to be

atraumatic. Osteoporosis poses a serious problem to pub-

examined can be determined. Bone diseases that might alter

lic health because of its prevalence and the cost associated

the shape or density of the bone, such as osteopetrosis or

with its comorbidity. According to the National Osteoporosis

ankylosing spondylitis, need to be ruled out. Previous frac-

tures or joint replacements that might alter the planning

must also be ruled out. Contraindications to DXA include

Table 2 Indications for bone densitometry.

pregnancy, recent (<5 days) oral administration of a con-

Females Older than 65 years trast agent, and a recent (<2 days) isotopic study. The

Younger than 65 years (postmenopausal patient does not require any speciﬁc preparation except for

or perimenopausal) the removal of any metal items located on the body area to

Males Older than 70 years be imaged.

Younger than 70 years with risk factors

of fracture

Areas of study

Both sexes Unexplained fracture

For most adult patients, the DXA examination should include

Diseases or chronic treatments

the lumbar spine and proximal femur; the forearm can also

Any patient to whom possible treatment

be included when hip or spine cannot be measured. In

is considered and to monitor the

children and adolescents (younger than 20 years), the mea-

response to treatment

27,28

surement is only performed in the lumbar spine. The

Document downloaded from http://www.elsevier.es, day 22/01/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.

Dual energy X-ray absorptimetry: Fundamentals, methodology, and clinical applications 413

Table 3 Parameters evaluated in dual energy X-ray absorptiometry.

BMC Bone mineral content

BMD Bone mineral density

SD Standard deviation

T-score Difference in number of standard deviations between the mean BMD value of the

patient and the mean of a young adult reference population of the same sex

Z-score Difference in number of standard deviations between the mean BMD value of the

patient and the mean of a reference population of the same race, sex and age

Osteopenia T-score between −1 and −2.5

Osteoporosis T-score ≤−2.5

BMI Body mass index

A/G ratio Ratio of android and gynoid A/G pelvic fat

ﬁnal result of the densitometry would be the lowest value

of the two regions studied.

The postero-anterior (PA) scan of the lumbar spine

includes the vertebral bodies of L1---L4, from which a mean

BMD of these four vertebrae is obtained. Vertebrae with

changes due to fracture or focal lesions will be excluded. To

assess the lumbar spine, at least two evaluable vertebrae

are required.

The study of the femur can indistinctly be performed on

the right or left hip, although it is useful to become used

to studying always the same side. Hips with changes due

to fracture, focal lesions or replacements will be excluded.

The total proximal femur or the femoral neck are used,

whichever is lowest.

The study of the non-dominant forearm is performed

when hip or spine cannot be measured (in order to have

a second measurable region), to obese patients (to bypass

Figure 1 Dual energy X-ray absorptiometry. PA study of the

technical difﬁculties), and to patients suffering from hyper-

lumbar spine. Patient’s positioning.

parathyroidism (since forearm bones change before the axial

skeleton).

scanning arm, measurements can be done with the patient

Patient positioning lying supine, whereas in systems with stationary arm mea-

Optimization of the patient’s position on the DXA table is surements can only be done in lateral decubitus.

essential. Incorrect positioning is one of the main causes of Regarding the scanning time, the ﬁrst densitometers with

errors in the estimation of BMD. In the PA study of the lum- pencil-beam scans took around 5 min per scan, but now the

bar spine, the patient lies supine on the table with the legs images can be obtained in less than one minute. According

ﬂexed over a support pad that reduces the lumbar lordosis

and approaches the spine to the table (Fig. 1).

For the study of the hip, the patient lies supine with legs

slightly in abduction in order to maintain the femoral axis

◦

straight and in internal rotation (15---30 ), in a way that the

lesser trochanter is not visible on the image (Fig. 2).

In the study of the forearm, the patient is seated, side

against the table with the arm resting on the tabletop with

the palm of the hand downwards patients sit beside the

table and with their forearm resting on it, with the hand

in pronation and strapped (Fig. 3).

The ﬁeld of view must include 1 --- 2 cm above and below

the area to be imaged. The bone must be straight and centered.

Acquisition

For most patients, the DXA includes PA projections of the

lumbar spine and proximal femur. The lateral spine is

not used in a standard osteoporosis study, but it is used Figure 2 Dual energy X-ray absorptiometry of the hip.

in vertebral morphometry. In those devices with mobile Patient’s positioning.

Document downloaded from http://www.elsevier.es, day 22/01/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.

414 R.M. Lorente Ramos et al.

DXA uses low radiation doses and generally most of these

devices do not require lead shielding of the room or special

protection measures for the technicians. In the DXA units

using pencil-beam scans, the equivalent surface dose for

spine and hip examinations is approximately 20---100 ␮Sv per

examination, and the equivalent effective dose is 1 --- 5 ␮Sv

per examination. For fan beam DXA, the dose is slightly

higher, approximately 56 ␮Sv for the hip, 59 ␮Sv for the

␮

spine, and 75 Sv for whole body. The disperse radiation that

the technician would receive being 1 m away from the table

using a fan beam unit, performing 4 hip studies and 4 spine

ones per hour would be of around 4 ␮Sv.

Analysis

Once the image is acquired, the assessment is carried

out by selecting different regions of interest (ROI). Inad-

equate placement of the ROIs is another important source

of errors. Although the equipment automatically proposes

speciﬁc areas, both the technician and the radiologist must

validate them and, given the case, rectify them.

In the study of the lumbar spine, the ROIs are placed

on the L1---L4 vertebral bodies (Fig. 4). It must be remem-

bered that D12 is usually the last vertebra connected to a

rib (although this is not always the case), and that L3 usually

has the longest transverse process.

In the study of the hip, the ROI must be placed at the

femoral neck, avoiding superimposition of the ischiopubic

ramus and the greater trochanter (Fig. 5). The system meas-

ures automatically the inclination of the femoral axis and

the rest of the ROIs.

In the forearm, the area of analysis is set at the distal

radius, with the line of reference at the urnal styloid process

Figure 3 Dual energy X-ray absorptiometry of the forearm.

(Fig. 6).

Patient’s positioning.

Screening of the causes of error

The radiologist must try to detect all those artifacts that

to the examination guidelines of the SERAM (Spanish Society might be possible causes of error in the acquisition, analysis

of Medical Radiology), the time of room occupancy for a and interpretation of the study (Table 4). Once the possible

spine or hip densitometry is 8 min, 10 min for the whole body, sources of error are identiﬁed, they must be excluded when

and the radiologist reporting time is 5 min. performing the analysis and subsequent interpretation.

PA Spine Bone Density Densitometric reference: L1-L4 BMD (g/m²) YA T-Score 1.42 2 Normal 1.30 1 1.18 0 L1 1.06 -1

0.94 Osteopenia -2 L2 0.82 -3 0.70 -4 Osteoporosis 0.58 -5 L3 20 30 40 50 60 70 80 90 100 Age (years) 1 2 2 BMD Young-Adult Young-Adult L4 Region (g/m²) (%) T-Score L1 0.987 87 -1.2 L2 0.930 78 -2.2 L3 1.053 88 -1.2 L4 0.961 80 -2.0 L1-L4 0.983 83 -1.6

L2-L4 0.982 82 -1.8

Figure 4 Dual energy X-ray absorptiometry image of a PA lumbar spine. The study includes the L1---L4 vertebral bodies.

Document downloaded from http://www.elsevier.es, day 22/01/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.

Dual energy X-ray absorptimetry: Fundamentals, methodology, and clinical applications 415 Left Femur Bone Density Densitometric reference: Neck BMD (g/m²) YA T-Score 1.22 2 Femoral Normal Neck 1.10 1 0.98 0 0.86 -1

0.74 Osteopenia -2 0.62 -3 0.50 -4 Osteoporosis 0.38 -5 20 30 40 50 60 70 80 90 100

Age (years)

1 2 3 Total Femur BMD Young-Adult Adj. to age Region (g/m²) (%) T-score (%) Z-score Neck 0.992 101 0.1 112 0.9 Ward’s 0.811 89 -0.8 110 0.6 Troc. 0.873 111 0.8 118 1.2 Diaphysis 1.331 - - - -

Total 1.095 110 0.8 117 1.3

Figure 5 Dual energy X-ray absorptiometry of the left hip. The lesser trochanter must not be visualized. The area of analysis

(ROI) is placed on the femoral neck.

UD Cubitus UD Radius

1 2 3 BMD Young adult Adj. to age Region (g/m²) (%) T-score (%) Z-score 33% UD radius 0.164 0.164 35 -6.7 52 -3.4 cubitus UD cubitus 0.131 0.131 - - - - 33% radius 0.378 0.378 43 -5.7 64 -2.4 33% cubitus 0.419 0.419 - - - - Both UD 0.152 0.152 - - - - Both 33% 0.398 0.398 - - - - 33% cubitus 33% radius Total radius 0.276 0.276 40 -6.7 58 -3.3 Total cubitus0.267 0.267 - - - -

Both total 0.272 0.272 - - - -

Figure 6 Dual energy X-ray absorptiometry of the forearm. The line of reference is placed at the urnal styloid.

Table 4 Sources of error in densitometry.

Technique Patient’s positioning

Movement

Region of interest (ROI) placement

Artifacts Foreign bodies Surgical material Calciﬁcations

Contrast media

Bone diseases Spondyloarthrosis

Fractures

Lytic or sclerotic lesions

Document downloaded from http://www.elsevier.es, day 22/01/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.

416 R.M. Lorente Ramos et al.

Figure 7 On the left, dual energy X-ray absorptiometry of the spine. Dense image superimposed on the L3 vertebral body (arrow).

On the right, the plain abdominal radiograph shows a calciﬁed mesenteric lymph node (arrow).

First, the correct positioning of the patient and absence islets, Paget’s disease, or hemangiomas. In cases of unknown

of movement during the scanning must be checked. Artifacts suspected bone lesion, complementary radiological studies

caused by superimposition of dense structures (including are required. It must be remembered that in patients with

surgical material, calciﬁcations or contrast agents) must be severe osteoporosis, the DXA scan might simulate a lytic

34,35 36

ruled out (Fig. 7). Bone lesions may also alter the anal- lesion (Fig. 11).

ysis and must be mentioned in the report.

The condition that most frequently distorts the anal- Parameters evaluated in bone densitometry

ysis is spondyloarthrosis, which is associated with bone Following the acquisition and analysis of DXA, the system

proliferation (osteophytes, endplate sclerosis) and morpho- calculates various parameters (Table 3). The BMC is the

logic changes (Fig. 8). Vertebral fractures (Fig. 9), lytic or quantity of calcium estimated by the energy absorbed by

sclerotic lesions (Fig. 10), metastases, lymphomas, bone it in a speciﬁc region. The BMD, much more relevant, is

PA Spine Bone Density Densitometric reference: L1-L4 BMD (g/m²) YAT-Score 1.54 3 Normal 1.42 2 1.30 1 L1 1.18 0 1.06 -1 0.94 Osteopenia -2 L2 0.82 -3 0.70 -4 Osteoporosis 0.58 -5 20 30 40 50 60 70 80 90 100

L3 Age (years)

1 2 3 BMD Young adult Adj. to age L4 Region (g/m²) (%) T-score (%) Z-score L1 0.883 78 -2.1 80 -1.8 L2 0.991 83 -1.7 85 -1.5 L3 1.326 111 1.1 113 1.3 L4 1.217 101 0.1 104 0.4 L1-L4 1.122 95 -0.5 97 -0.3

L2-L4 1.184 99 -0.1 101 0.1

Figure 8 Dual energy X-ray absorptiometry image of spine. The increase of density in the radiography (arrow) corresponds to

an L3 and L4 sclerosis. In the analysis, the bone mineral density (BMD) in L3 and L4 is higher than in L1 and L2. Exclusion of both

vertebrae shows a T-score L1---L2 = −1.9 (L1---L4 = −0.5).

Document downloaded from http://www.elsevier.es, day 22/01/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.

Dual energy X-ray absorptimetry: Fundamentals, methodology, and clinical applications 417

BMD (g/m²) YAT-Score 1.48 2

1.36 1

1.24 0

[L1] 1.12 -1

1.00 -2

[L2] 0.88 -3

0.76 -4 L3 0.64 -5 20 30 40 50 60 70 80 90 100 Age (years) L4 Densitometry

BMD BMC Area YA YA AM AM

Region (g/m²) (g) (m²) (Z) T-score (Z) Z-score L1 0.884 9.20 10.23 77 -2.2 83 -1.5 L2 1.054 11.12 10.55 85 -1.5 91 -0.8 L3 0.876 10.28 11.74 71 -3.0 76 -2.3 L4 0.844 11.58 13.73 68 -3.3 73 -2.6

L3-L4 0.858 21.87 25.47 69 -32 74 -2.5

Figure 9 Above, dual energy X-ray absorptiometry of spine. Increased density of the L2 vertebral body (arrow) that corresponds

to a fracture (arrow) on the radiograph below. Both images show a calciﬁcation (arrowheads) that is not superimposed to the spine

(not affecting the analysis). PA Spine Bone Density Densitometric reference: L2-L4 BMD (g/m²) YAT-score 1.44 2 Normal 1.32 1 1.20 0 1.08 -1

0.96 Osteopenia -2 0.84 -3 R 0.72 -4 Osteoporosis (L1) 0.60 -5 20 30 40 50 60 70 80 90 100 Age (years) L2 123 BMD Young-Adult Adj. to age Region (g/m²) (%) T-score (%) Z-score L3 L1 8.822 73 -2.6 93 -0.5 L2 0.751 63 -3.7 79 -1.7 L3 0.851 71 -2.9 90 -0.8 L4 L4 0.920 77 -2.3 97 -0.2

L2-L4 0.847 71 -2.9 69 -0.9

Figure 10 Dual energy X-ray absorptiometry and AP radiograph of spine (right). Increased density in the pedicle of right L1 (arrow)

corresponding in the radiography to a dense pedicle. The vertebra must be excluded from the analysis.

Document downloaded from http://www.elsevier.es, day 22/01/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.

418 R.M. Lorente Ramos et al.

Figure 12 Whole body dual energy X-ray absorptiometry.

Patient’s positioning.

the mean quantity of mineral per unit area. It is calculated

dividing the BMC by unit area (g/cm ). The values used for

diagnosis (T- and Z-score) are obtained by comparison with

the reference database.

The T-score is the value used to diagnose osteoporosis

in postmenopausal women and in men aged 50 and over.

T-score is the number of standard deviations the patient’s

BMD above or below the mean for the young adult reference

population of the same sex. A T-score >−1.0 is considered

normal, osteopenia when the T-score is between −1 and

−

2.5, and osteoporosis is a T-score <−2.5.

The Z-score is used in premenopausal women, in men

younger than 50 years, and in children and adolescents (up

Figure 11 Dual energy X-ray absorptiometry of spine in a

to 20 years). Z score is the number of standard deviations the

patient with severe osteoporosis. The image simulates a defect

patient’s BMD above or below the mean for the reference

of L3 and L4 (arrows), not conﬁrmed by the radiographies.

population of the same race, sex and age. A Z-score <−2

World Health Organization BMI Classification BMI=16.5 (kg/m2)

13 18.5 25 30 35

Underweight Normal Overweight Obese

35 50 68 82 95 Weight (kg) for height=165.0 cm Composition Tissue Total mass Region Tissue Fat Lean BMC Fat lib Centile Region (% fat) (kg) (%fat) (g) (g) (g) (g) Left leg 21.7 - 8.1 20.6 7.695 1.673 6.021 427 6.4 Left torso 15.1 - 10.6 14.6 10.236 1.547 8.689 380 9.0 Total left 16.6 - 22.5 15.8 21.347 1.543 17.804 1.146 18.9 Right arm 8.2 - 2.2 7.6 2.065 169 1.896 151 2.0 Right leg 21.7 - 8.3 20.5 7.843 1.703 6.140 442 6.5 Right torso 15.1 - 10.0 14.6 9.667 1.463 8.203 360 8.5 Total right 16.5 - 23.0 15.6 21.703 3.583 18.120 1.276 19.3 Arms 8.2 - 4.5 7.6 4.229 345 3.884 304 4.1 Legs 21.7 - 16.4 20.6 15.537 3.376 12.161 870 13.0 Torso 15.1 - 20.6 14.6 19.902 3.010 16.892 740 17.6

Android 10.8 - 2.5 10.6 2.432 263 2.168 50 2.2

Gynoid 28.0 - 7.0 27.1 6.822 1.910 4.911 225 5.1

Total 16.6 - 45.5 15.7 43.049 7.125 35.924 2.422 38.3

Figure 13 Whole body dual energy X-ray absorptiometry. Underweight patient.

Document downloaded from http://www.elsevier.es, day 22/01/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.

Dual energy X-ray absorptimetry: Fundamentals, methodology, and clinical applications 419

standard deviations is deﬁned as ‘‘below the expected range 38

for age’’.

Reporting

The report should be tailored to the particular char-

acteristics of each center, but in general it must

comply with some minimum requirements. In addition to

the patient’s personal details, the report must include

the date of examination, the type of DXA system used,

as well as the protocol followed. It must also specify if

any region of analysis has been excluded and the rea-

sons, and the possible presence of artifacts or suspected

lesions.

The report must also include the scans obtained for each

region of study, with a graph showing the patient’s data and

then related back to the reference curve, the numerical val-

ues of the BMD, T- and Z-scores, and ﬁnally, a diagnostic

category based on WHO criteria. The value considered most

useful to diagnose a particular patient must be speciﬁed.

In some cases, an estimation of the risk of fracture is also

included. In general, the risk of fracture doubles for each

standard deviation reduction in BMD. The study’s conclusion

should be global, taking into account the lowest value of the

areas studied. In patients with previous studies, any possible

changes should be reported.

Estimation of the body composition using dual

energy X-ray absorptiometry

Several techniques have been used to assess the composition

and quantiﬁcation of the body fat, mostly anthropometric

methods such as the waist circumference, the waist-to-hip

ratio, and the skin fold thickness. Among the imaging stud-

ies, CT stands out to measure visceral fat, but with some

drawbacks such as radiation, among others. Bioimpedance

techniques and isotope dilution have also been

used.40

Whole body DXA allows for an easy and rapid mea-

surement of body composition. It provides an estimate of

body fat, and when needed, it may also estimate the BMD

of the whole body. DXA is very accurate, with a margin of

error of 2---6% for body composition. Unlike anthropomet-

ric methods, DXA has advantage of providing regional and

whole-body measurements. DXA is increasingly being used;

for many clinicians it is a common tool and for some authors

it is the reference standard.

Indications

Whole body DXA is used in eating disorders, especially those

where there might be a hormonal disorder or cardiovascular

risk factors. It is also used in gastrointestinal diseases, hepa-

tobiliary disorders, advanced renal failure, in endocrinal

diseases, bone disorders such as Paget’s disease or osteopet-

rosis, lung diseases and various chronic treatments. DXA

may also help design the diet in patients with mal-

nutrition and in the follow-up of patients with eating

disorders.

Figure 14 Whole body dual energy X-ray absorptiometry.

Obese patient.

Methodology

Patient’s positioning. The patient lies supine, centered on

the table with the arms along sides of the body, hands

Document downloaded from http://www.elsevier.es, day 22/01/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.

420 R.M. Lorente Ramos et al.

facing the legs without touching them and the thumbs up The system places the ROIs automatically. The technician

(Fig. 12). If the patient is wider than the width of the revises the ROIs and, if necessary, modiﬁes them, although

DXA table, a half-body scanning is performed (including it is advisable to manipulate them as little as possible.

the neck and head, one whole side with their correspond- The changes performed on an image are automatically

ing arm and leg). In this case, the patient lies supine introduced on the other. The ROIs correspond to anatomic

but is positioned off the center line of the table to regions: the head up under the chin; arms separated from

ensure that one side is completely included in the scan the body with the cut lines passing through the armpits; fore-

ﬁeld. arms separated from the body, legs separated from the arms

Analysis. As in BMD studies, the correct positioning and and the medial cut lines located between both legs; the lim-

absence of motion artifacts must be veriﬁed. Following the its of the spine are adjacent to it, on both sides; pelvis: the

acquisition, two images of the whole body are obtained, upper cut line located immediately above the pelvis and cut

one of bones and other of soft tissues (Figs. 13 and 14). lines through the femoral neck without touching the pelvis.

World Health Organization BMI Classification BMI = 57.0

13 18.5 25 30 60 Underweight Normal Overweight Obese

36 51 69 83 165

Weight (kg) for height = 166.0 cm

Figure 15 Whole body dual energy X-ray absorptiometry. The patient is wider than the width of the table so the system performs

half-body scanning (left). The left side is completely included in the scan ﬁeld, and a contralateral estimation is performed.

Document downloaded from http://www.elsevier.es, day 22/01/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.

Dual energy X-ray absorptimetry: Fundamentals, methodology, and clinical applications 421

In patients wider than the width of the scan ﬁeld where only

a half-body is scanned, the ROI is placed as explained and

the system provides a total estimation (Fig. 15).

Interpretation. The system provides various parameters

(Table 3) such as the body mass index (BMI), quantiﬁcation

of body fat, distribution of the abdominal fat, or values

obtained from these data such as the android/gynoid ratio

(A/G ratio) in the abdominal fat.

The BMI is an anthropometric indicator designed for adult

males and non-pregnant females, but it does not distinguish

between fat and muscle and therefore cannot be assessed

in athletes. The BMI is a number calculated from a person’s

weight and height and is measured in kg/m . The distri-

bution of body tissues is expressed as percentages for the

whole body and per regions: percentage of fat (fat mass),

of soft parts and muscle (lean mass) and of body (BMD).

Several studies have calculated curves that can be used as

41---44

a reference for different populations.

In addition to the composition by anatomic regions, it

calculates the distribution of fat in predeﬁned regions in the

pelvic area: android (central, the lower limit is the pelvis

and the lateral limit is the arms) and gynoid (hip and thighs,

the lateral limits are the outer part of the legs) (Fig. 16).

The android/gynoid (A/G) ratio of the fat is the relation

between the fat percentage of the android and the gynoid

region. The excess of abdominal fat (android) is associated

with a number of cardiovascular risk factors.

The determination of the A/G ratio using DXA is an easy

and useful tool to assess the distribution of pelvic fat. This

ratio may have a role in evaluating the cardiovascular risk

in overweight or underweight patients.

BMD values of the whole body are useful to estimate min-

eralization, but not to diagnose osteoporosis (it is necessary

to study the spine and hip in order to compare the results

with the reference curves to reach a diagnosis).

Other applications of dual energy X-ray

absorptiometry

Over the time, other potential clinical applications using

DXA are being proposed. DXA systems now offer the pos-

sibility of creating personalized analysis areas to perform

composition measurements. There are also studies in the

ﬁeld of orthopedics that investigate implant integration by

assessing the regional mineralization. Moreover, its use

in lipodystrophy and lipoatrophy is being studied, especially

in HIV patients.

Conclusion

DXA is a fast and reliable technique with little radiation

exposure. It is the technique of choice in the diagnosis and

monitoring of osteoporosis since it objectively measures the

most relevant parameters. Moreover, it is useful to mea-

Figure 16 Android (continuous line) and gynoid (dotted line) sure the whole body composition and its distribution by

areas of fat. regions. There are other less common applications in poten-

tial expansion. Knowledge of the technique, its indications,

methodology and applications are key issues to optimize its

results and streamline its use.

Document downloaded from http://www.elsevier.es, day 22/01/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.

422 R.M. Lorente Ramos et al.

Authorship interpretation of peripheral X-ray absorptiometry examina-

tions. Osteoporosis Int. 2005;16:2149---56.

17. Picard D, Brown JP, Rosenthall L, Couturier M, Lévesque J,

1 Responsible for the integrity of the study: RMLR.

Dumont M, et al. Ability of peripheral DXA measurement to

2 Conception of the study: RMLR.

diagnose osteoporosis as assessed by central DXA measurement.

3 Design of the study: RMLR and JAA.

J Clin Densitom. 2004;7:111---8.

4 Acquisition of data: RMLR, JAA and NAG.

18. Albanese CV, Diessel E, Genant HK. Clinical applications of

5 Analysis and interpretation of data: RMLR, AMH, JMGG body composition measurements using DXA. J Clin Densitom.

and JGM. 2003;6:75---85.

6 Statistical analysis: RMLR 19. Baim S, Binkley N, Bilezikian JP, Kendler DL, Hans DB,

7 Bibliographic search: RMLR and JAA. Lewiecki EM, et al. Ofﬁcial position of the International Soci-

ety for Clinical Densitometry and executive summary of the

8 Drafting of the paper: RMLR.

2007 ISCD position development conference. J Clin Densitom.

9 Critical review with intellectually relevant contrib-

2008;11:75---91.

utions: RMLR, JAA, NAG, AMH, JMGG and JGM.

20. Lewiecki EM, Watts NB, McClung MR, Petak SM, Bachrach LK,

10 Approval of the ﬁnal version: RMLR, JAA, NAG, AMH,

Shepherd JA, et al. Ofﬁcial positions of the International

JMGG and JGM.

Society for Clinical Densitometry. J Clin Endocrinol Metab.

2004;89:3651---5.

21. Consensus Development Conference. Diagnosis, prophylaxis and

Conﬂicts of interest

treatment of osteoporosis. Am J Med. 1993;94:646---50.

22. National Osteoporosis Foundation. Available from:

The authors declare not having any conﬂicts of interest. http://www.nof.org/ [accessed 06.05.11].

23. Internacional Osteoporosis Fundation (IOF). Osteoporosis in

the European Community: a call to action. An audit of pol-

References icy development since 1998. Brussels: IOF; 2001. Available in:

http://www.iofbonehealth.org [accessed 06.05.11].

1. Cummings SR, Bates D, Black DM. Clinical use of bone densito- 24. Álvarez Sanz C, Rapado A, Díaz Curiel M, García JJ, Carrasco

metry. JAMA. 2002;288:1889---97. JL, Honorato J, et al. Prevalencia de osteoporosis determinada

2. Bachrach LK. Dual energy X-ray absorptiometry (DEXA) mea- por densitometría en la población femenina espanola.˜ Med Clin.

surements of bone density and body composition: promise and 2001;116:86---8.

pitfalls. J Pediatr Endocrinol Metab. 2000;13:983---8. 25. Assessment of fracture risk and its application to screening for

3. Yang Y, Zhu WD, Paton NI. Comparison of dual-energy X- postmenopausal osteoporosis. Report of a WHO Study Group.

ray absorptiometry machines for measuring fat distribution World Health Organ Tech Rep Ser. 1994;843:1---129.

changes of HIV-associated lipodystrophy. Antiviral Therapy. 26. Lenchik L, Rochmis P, Sartoris DJ. Optimized interpretation

2004;9:771---8. and reporting of dual X-ray absorptiometry (DXA) scans. Am J

4. Spittlehouse AJ, Smith TW, Eastell R. Bone loss around 2 differ- Roentgenol. 1998;171:1509---20.

ent types of hip prostheses. J Arthroplasty. 1998;13:422---7. 27. Baim S, Leonard MB, Bianchi ML, Hans DB, Kalkwarf HJ, Lang-

5. Wiklund P, Toss F, Weinehall L, Hallmans G, Franks PW, Nord- man CB, et al. Ofﬁcial positions of the International Society for

ström A, et al. Abdominal and gynoid fat mass are associated Clinical Densitometry and executive summary of the 2007 ISCD

with cardiovascular risk factors in men and women. J Clin pediatric position development conference. J Clin Densitom.

Endocrinol Metab. 2008;93:4360---6. 2008;11:6---21.

6. Adams JE. Single and dual energy X-ray absorptiometry. Eur 28. Rauch F, Plotkin H, DiMeglio L, Engelbert RH, Henderson RC,

Radiol. 1997;7:S20---31. Munns C, et al. Fracture prediction and the deﬁnition of osteo-

7. Genant HK. Current state of bone densitometry for osteoporo- porosis in children and adolescents: the ISCD 2007 pediatric

sis. Radiographics. 1998;18:913---8. ofﬁcial positions. J Clin Densitom. 2008;11:22---8.

8. Alcaraz Mexía MJ. La densitometría ósea. In: Tardáguila Montero 29. Lentle BC, Prior JC. Osteoporosis: what a clinician expects to

F, del Cura Rodríguez JL, editors. Radiología ortopédica y radi- learn from a patient’s bone densitometry examination. Radiol-

ología dental. Una guía práctica. Buenos Aires, Madrid: Médica ogy. 2003;228:620---8.

Panamericana; 2004. p. 91---8. 30. Catálogo de exploraciones de la Sociedad Espanola˜ de Radi-

9. Blake GM, Fogelman I. Technical principles of dual energy X-ray ología Médica (SERAM). http://www.SERAM.es; 2009 [accessed

absorptiometry. Semin Nucl Med. 1997;27:210---28. online 29.04.11].

10. Genant HK, Grampp S, Glüer CC, Faulkner KG, Jergas M, Enkelge 31. Kalender WA. Effective dose values in bone mineral measure-

K, et al. Universal Standardization for dual-X-ray absorptiom- ments by photon absorptiometry and computed tomography.

etry: patient and phantom cross-calibration results. J Bone Osteoporosis Int. 1992;2:82---7.

Mineral Res. 1994;9:1503---14. 32. Steel SA, Baker AJ, Saunderson JR. An assessment of the radia-

11. Dasher LG, Newton CD, Lenchik L. Dual X-ray absorptiometry in tion dose to patients and staff from a Lunar Expert-XL fan beam

today’s clinical practice. Radiol Clin North Am. 2010;48:541---60. densitometer. Physiol Meas. 1998;19:17---26.

12. Bauer JS, Link TM. Advances in osteoporotic imaging. Eur J 33. Staron RB, Greenspan R, Millar TT, Bilezikian JP, Shane

Radiol. 2009;71:440---9. E, Haramati N. Computerized bone densitometric analysis:

13. Adams JE. Quantitative computed tomography. Eur J Radiol. operator-dependent errors. Radiology. 1999;211:467---70.

2009;71:415---24. 34. Jacobson JA, Jamadar DA, Hayes CW. Dual X-ray absorptiometry.

14. Lang T, Augat P, Majumdar S, Ouyang X, Genant HK. Non- Recognizing image artifacts and pathology. Am J Roentgenol.

invasive assessment of bone density and structure using 2000;174:1699---705.

computed tomography and magnetic resonance. Bone. 1998;22 35. Lorente-Ramos R, Azpeitia-Armán J, Munoz-Hernández˜

Suppl:S149---53. A, García-Gómez JM, Díez-Martínez P, Grande-Bárez

15. Chun KJ. Bone densitometry. Semin Nucl Med. 2011;41:220---8. M. Dual-energy X-ray absorptiometry in the diagnosis of

16. Blake GM, Chinn DJ, Steel SA, Patel R, Panayiotou E, Thorpe osteoporosis. A practical how-to guide. Am J Roentgenol.

J, et al. A list of device-speciﬁc thresholds for the clinical 2011;196:897---904.

Document downloaded from http://www.elsevier.es, day 22/01/2015. This copy is for personal use. Any transmission of this document by any media or format is strictly prohibited.

Dual energy X-ray absorptimetry: Fundamentals, methodology, and clinical applications 423

36. Theodorou DJ, Theodorou SJ. Dual-energy X-ray absorptiome- content and areal bone mineral density in children aged

try in clinical practice: application and interpretation of scans 8---18: the Fels longitudinal study. Am J Clin Nutr. 1998;68:

beyond the numbers. Clin Imaging. 2002;26:43---9. 1111---7.

37. Wahner H. Technical aspects and clinical interpretation of 42. Coin A, Sergi G, Minicuci N, Gianni S, Barbiero E, Manzato E,

bone mineral measurements. Public Health Rep. 1989;104 et al. Fat-free mass and fat mass referente values by dual-

Suppl:S27---30. energy X-ray absorptiometry (DEXA) in a 20---80 year-old italian

38. Gordon CM, Bachrach LK, Carpenter TO, Crabtree N, Fulei- population. Clin Nutr. 2008;27:87---94.

han GE, Kutilek S, et al. Dual energy X-ray absorptiometry 43. Demura S, Sato S. Non-linear relationships between visceral

interpretation and reporting in children and adolescents: the fat area and percent regional fat mass in the trunk and the

2007 ISCD pediatric ofﬁcial positions. J Clin Densitom. 2008;11: lower limbs in Japanese adults. Eur J Clin Nutr. 2008;62:

43---58. 1395---404.

39. Tzotzas T, Krassas GE, Doumas A. Body composition analysis in 44. Bozkirli E, Ertorer ME, Bakiner O, Tutuncu NB, Demirag NG. The

obesity: radionuclide and non radionuclide methods. Hell J Nucl validity of the World Health Organisation’s obesity body-mass

Med. 2008;11:63---71. index criteria in a Turkish population: a hospital-based study.

40. Fogelholm M, van Marken Lichtenbelt W. Comparison of body Asia Pac J Clin Nutr. 2007;16:443---7.

composition methods: a literature analysis. Eur J Clin Nutr. 45. Mahabir S, Ettinger S, Johnson L, Baer DJ, Clevidence BA, Hart-

1997;51:495---503. man TJ, et al. Measures of adiposity and body fat distribution

41. Maynard LM, Chumlea WC, Roche AF, Wisemandle WA, Zeller in relation to serum folate levels in postmenopauseal women in

CM, Towne B, et al. Total-body and regional bone mineral a feeding study. Eur J Clin Nutr. 2008;62:644---50.