Imaging in Pediatric Rheumatic Diseases

Radiography

Radiography is important in the evaluation of bone and joint abnormalities, particularly for assessment of osteochondral changes, growth disturbances, soft tissue calcifications, and osteopenia. Despite recent advances in X-ray technology, however, soft tissue changes are less evaluable. In general, the radiation dose is low, with an effective dose ranging from around 0.001 mSv for a hand radiograph to 0.2 mSv for a full spine radiograph, compared with the normal background radiation of around 1.5 to 5 mSv per year. ^, Nevertheless, care should be taken to minimize the total radiation burden, particularly to children with chronic diseases, who may be subject to multiple investigations over a lifetime.

Special attention to imaging technique is paramount for correct diagnosis. For example, mal positioning, with a slightly flexed right and an extended left knee joint will typically show an apparent enlarged right epiphysis, suggestive of overgrowth resulting from persistent joint inflammation. Likewise, low-resolution images may fail to demonstrate both true erosions and osteopenia, underscoring the need for age- and disease-specific protocols in musculoskeletal imaging.

Sonography

Recent advances in sonography, including technical progress and the development of a better evidence base, have stimulated the increased use of this modality in the assessment of pediatric joint disease. ^, Sonography is ideal for assessing the pediatric musculoskeletal system, largely because of its ability to visualize the various periarticular and intraarticular structures including cartilage and bone without the need for radiation ( Fig. 9.1 ). Sonography is very sensitive in detecting synovitis, tenosynovitis, paratenonitis, and enthesitis ( Fig. 9.2 ) Multiple joints can be assessed at the same time, and even very young children can be examined without sedation. It can also be used to guide joint aspiration or injection ( Fig. 9.3 ). Intraarticular masses may also be detected with sonography, although their appearance is often nonspecific. Sonography has the highest spatial resolution of the commonly used imaging techniques, allowing for a detailed view of even very small structures such as the pulleys of the flexor tendons or entheses of the extensor tendons and collateral ligaments of the fingers ( Figs. 9.4, 9.5, and 9.6 ). Sonography can also be used to assess other periarticular soft tissue structures such as synovial cysts.

All these structural findings are assessed using B (brightness)-mode sonography. This technique uses the transmission and reflection of high-frequency longitudinal mechanical waves (ultrasonic waves). The image is generated by the differing energy of waves reflected from surfaces of different tissues. The reflected “echoing” waves are analyzed and displayed by a computer, which creates a real-time display of tissues and structures being examined. The intensity of the brightness indicates the energy of the reflected sound waves. The resulting image of black, white, and gray pixels is described as follows: hyperechoic (white), intermediate echo (gray), hypoechoic (gray/dark gray), or anechoic (black). The term isoechoic means that the visualized tissue is of the same echogenicity as surrounding tissues.

Vascular anatomy can be assessed by combining B-mode sonography with Doppler sonography, which will detect blood flow. Synovial hyperemia, for example, leads to increased Doppler signals because of the increased blood flow. Both power and color Doppler sonography can be used. In B-mode it is important to use the highest frequency possible for good resolution and the lowest frequency necessary for good penetration. The frequency may have to be adjusted for different views, the patient’s age, and body mass index. In Doppler sonography it is also important to adjust the frequency according to the same principles as in B-mode but also to use low flow settings including a low pulse repetition frequency of less than 1 kHz (preferably around 0.5 kHz) and low wall filters. The choice between color and power Doppler sonography is largely dependent on which of the two techniques performs better on any given machine after settings have been optimized.

The major limitation for widespread use of sonography has been a lack of standardization in the sonographic assessment of the pediatric joint, as well as the challenge of reliably distinguishing pathological findings from normal sonographic anatomy.

Computed Tomography

Multidetector computed tomography (CT) scanners generate detailed high-resolution images of bone and can be used to evaluate the joint space and detect adjacent bone abnormalities including tarsal coalitions, bone erosions, subchondral cysts, or primary osseous lesions, such as osteoid osteoma. Current-generation CT scanners are fast, and sedation is generally not required. Intravenous contrast may be helpful for soft tissue assessment. Because magnetic resonance imaging (MRI) provides better soft tissue contrast without need for radiation, CT is primarily used to provide detailed assessment of osseous structures in locations that are difficult to assess with conventional radiography ( Fig. 9.7 ). The effective radiation doses are higher than those for conventional radiography, varying from 1 to 10 mSv depending on the size of the body part examined, the type of procedure, and the type of CT equipment and its operation. Focused, age-specific protocols should be employed.

Cone Beam Computed Tomography

Cone beam computed tomography (CBCT) is a feasible and accurate low-dose technique for assessment of the temporomandibular joint (small field of view) and for craniofacial morphometrics (large field of view) ( Fig. 9.8A and B ) . It uses a divergent X-ray, forming a cone centered on two-dimensional detectors, obtaining two-dimensional images, and reformatting the information obtained into a volume. The examination time is around 5 to 20 seconds, with radiation doses up to 15 times lower than that of conventional CT.

Magnetic Resonance Imaging

MRI uses a strong magnetic field and radiofrequency waves to generate images. For musculoskeletal imaging, field strengths of 1.5 to 3 tesla (T) are used; 3 T is preferred for detailed evaluation of smaller structures because it allows for greater spatial resolution. Although MRI provides exquisite images with high spatial resolution and excellent tissue contrast, high cost, limited availability, long scan time, and the need for sedation in children under the age of 4-6 years have limited its more widespread use. Nevertheless, MRI is an excellent modality to examine the musculoskeletal system (perhaps excepting cortical bone), including bone marrow, hyaline and fibrocartilage, ligaments, menisci, synovium, tendon sheaths and joint capsule, joint fluid, and the unossified cartilaginous skeleton.

MRI studies consist of a number of imaging series that typically are composed of different imaging planes and pulse sequences. A pulse sequence is defined by numerous parameters that affect tissue contrast and spatial resolution . Different sequences are grouped according to the dominant influence on the appearance of tissues: T1-weighted, T2-weighted (or water sensitive), diffusion-weighted, flow sensitive, and “miscellaneous.” T1- and T2-weighted sequences, both two-dimensional (2D) and three dimensional (3D), are most often acquired by spin echo (SE) or gradient recalled echo (GRE) technique.

On T1-weighted images, water components (from tissue edema, joint effusion, cerebrospinal fluid, etc.) appear as “dark” (low) signal and on T2-weighted images as “bright” (high) signal ( Table 9.2 ). In contrast to fluid, fat presents as “bright” signal on T1-weighted images and as “intermediate to bright” signal on T2-weighted images. Proton density (PD)–weighted sequences share features of both T1 and T2 and are helpful for differentiation between fluid (bright), hyaline cartilage (intermediate), and fibrocartilage (dark). Fat suppression is performed to suppress the bright signal from fat on T1, T2, or PD sequences. It is particularly useful for the assessment of pathologically increased contrast uptake on T1-weighted images and also for differentiation of various tissues. High-resolution 3D techniques using isometric voxels allowing for reconstructions in different planes are helpful for assessing complex joint anatomy.

Diffusion-weighted sequences (DWI or DW-MRI) use the diffusion of water molecules to generate contrast in MRIs. In general, highly cellular tissues or those with cellular swelling exhibit lower diffusion coefficients, for example, apparent diffusion coefficient (ADC) values. On the other hand, inflammatory tissue exhibits higher ADC values.

Descriptions of most MRI sequences refer to the shade of gray of tissues or fluid using the term intensity as high (white), intermediate (gray), or low signal intensity (black). These terms are relative, and it is therefore necessary to compare the signal intensity to surrounding tissues (e.g., muscle).

The most commonly used contrast agents in MRI are gadolinium-based contrast agents (GBCAs). In general, these agents do not enter intact cells but distribute in the interstitial/extracellular spaces within a few minutes after intravenous injection. T1 values are shortened wherever gadolinium accumulates, leading to a high signal on T1-weighted images. GBCAs have been considered safe during the past three decades of their clinical use, and the reported incidence of adverse reactions, both acute incidents and nephrogenic systemic fibrosis, are extremely low in children. More recently there has been a growing concern over the reported GBCA deposition in the body; however, there is no evidence for any clinical long-term adverse effects.

Both normal and diseased tissue will take up gadolinium contrast, and the degree of enhancement depends on when imaging occurs. In the early stage (1 to 3 minutes) gadolinium remains primarily in the blood pool and interstitial/extracellular spaces. In later phases there is diffusion of contrast into joint fluid, underscoring the need for standardized timing of postcontrast images, particularly when evaluating the effect of treatment.

Contrast can also be given as an intravenous infusion over a few minutes, so-called perfusion imaging . This technique may be helpful for quantification and monitoring of synovial inflammation, for characterizing ischemic or hyperemic areas, and for recognition of epiphyseal ischemia.

Biochemical MRI techniques (delayed gadolinium-enhanced MRI cartilage imaging [dGEMRIC], T2 and T1rho mapping, among others) allow for quantitative evaluation of cartilage matrix composition for early detection and monitoring of degenerative changes. These techniques are not widely used in daily practice.

Whole-body MRI (WB-MRI), using coronal fluid sensitive short tau inversion recovery (STIR) imaging, is used for assessment of multifocality. Its use in detecting asymptomatic lesions in chronic recurrent multifocal osteomyelitis is commonly used, but more knowledge of the spectrum of normal variations that may mimic pathology is needed.

To summarize, in clinical practice, traditional T1-weighted and T2-weighted sequences are very useful for evaluating anatomy such as cartilage, whereas fluid sensitive sequences such as T2 and PD with fat suppression help detect and differentiate effusion and edema. A cartilage-specific sequence is advised for more detailed evaluation of morphological changes ( Fig. 9.9 ), whereas a contrast-enhanced fat suppressed T1 series is helpful when further characterization of the synovium is needed. This may be the case in wrist involvement when the initial series shows numerous bony depressions and joint fluid pockets greater than 2 mm, and the differentiation between normal variations and true disease is sought. In general, a radiograph should be added when necessary to assist in the interpretation of the MRI examination.

Arthrography

Currently, arthrography is rarely indicated. Intraarticular contrast injection may, however, be combined with CT or (now more frequently) with MRI to better delineate joint detail including the evaluation of intraarticular loose bodies or labral tears within the shoulder or hip joint.

Bone Scintigraphy

A number of bone scintigraphy techniques have been used, including the radionuclide bone scan (planar focal or whole body images with additional localized or spot views), single photon emission tomography (SPECT) allowing for visualization of the 3D distribution of the radiopharmaceutical in the skeleton), SPECT/CT images consisting of a SPECT acquisition combined with CT, multiphase bone scan and positron emission tomography (PET) CT, or MRI with radionuclide bone scan as the cornerstone. Although highly sensitive to changes in local blood flow and bone remodeling activity at a relatively modest radiation dose, scintigraphy is generally not recommended in the routine management of rheumatic disease in children. Similar to MRI, a thorough knowledge of growth- and maturation-related bone changes reduces the risk of false-positive statements.

Radiography

Bone is a dynamic tissue that undergoes a continuous remodeling process throughout life. The number, shape, and structure of bones changes significantly during the maturation process, and there is a wide variety of normal variants that may simulate disease . From 0 to 3 months of age, a physiological periosteal reaction can be seen along the long tubular bones, not to be mistaken for inflammatory disease or trauma. Further, the long tubular bones typically evolve from a “stick-like” fetal shape with only a mild diaphyseal constriction or concavity to a well-defined constriction in childhood. Similarly, the shape of the vertebral bodies undergoes a significant change during the last trimester of gestation and continues through infancy and childhood. Somewhat flattened and dense vertebrae, vertebrae with a mild anterior hook or coronal clefts (in boys), would, for instance, be perceived as normal in a newborn baby, but would represent pathology in an older child.

In neonates and young children, radiography demonstrates wide apparent joint spaces representing immature unossified epiphyses. These chondroepiphyses will eventually ossify, reducing the apparent joint space to the thickness of the opposing layers of articular cartilage and any intervening joint fluid.

Occasionally, intraarticular gas can be seen as a normal finding, but it can also be seen after infection, trauma, or invasive procedure. Radiographically, normal intraarticular gas appears as an intraarticular crescentic lucency and is caused by sudden lowering of intraarticular pressure by muscle pulls or external traction. Intraarticular gas can also be identified with sonography or MRI. On MRI, intraarticular gas may simulate meniscal tears, intraarticular loose bodies, or chondrocalcinosis.

To date, references for radiographic-based appearances of the skeleton in children are lacking, perhaps except for the hand ( Table 9.3 ). Recent research has shown that the Greulich and Pyle hand atlas is applicable to most ethnicities.

Sonography

Normal Sonoanatomy

Bones usually serve as a reference point in musculoskeletal sonography. Whereas the skeleton of an adult is completely ossified except for the articular cartilage, a varying degree of hyaline cartilage and, in some locations, fibrocartilage is present in addition to articular cartilage in children ( Fig. 9.10 ). At birth, the primary ossification centers in the diaphysis of the long bones are already present, whereas the secondary ossification centers in the epiphysis, except for the distal femur, develop subsequently. In some locations, for example, the humeral head and multiple secondary ossifications centers give an irregular appearance that should not be misinterpreted as pathological ( Figs. 9.1 and 9.11 ). The same principle applies to the short bones with one or several ossification centers appearing over time. The degree of ossification increases with age but varies significantly within the same age group according to the individual stage of maturation (pubertal stages). Children of the same chronological age may be at different stages of skeletal maturation and therefore display different degrees of ossification. The degree of ossification and therefore the individual progress of maturation can be reliably imaged with sonography. In any given bone, significant variability of the progress of ossification can also be observed independent of the stage of maturation. Some basic knowledge of the progress of ossification in various anatomical areas is therefore helpful. A free online resource on musculoskeletal ultrasonography included summaries of ossification ( www.ped-mus.com ).

The presence of variable amounts of cartilage can present significant challenges for the interpretation of radiographs. In contrast, sonography can delineate the cartilage outline very well. Depending on the age of the child, anechoic or hypoechoic cartilage will define the bone contour in a joint, as opposed to the hyperechoic outline of the fully ossified bone seen in adults ( Fig. 9.12 ). A careful scanning technique is essential to ensure the clear differentiation of cartilage from the possible presence of fluid in the joint, which would also appear anechoic or hypoechoic. Cartilage is also not displaceable by the ultrasound probe, whereas fluid is displaceable. The anatomical location and shape of the anechoic area will help distinguish between fluid and cartilage, and the most important maneuver to distinguish cartilage from fluid may be simply moving the joint while scanning. Cartilage will keep the same shape whereas fluid will change shape with movement.

The growth plate is seen as an anechoic or hypoechoic line separating the epiphysis from the meta/diaphysis. The cartilage itself may display hyperechogenic spots, which represent vascular channels. ^, These are physiological in children and should not be interpreted as pathological ( Fig. 9.13 ). The Outcome Measures in Rheumatoid Arthritis Clinical Trials (OMERACT) Ultrasound Group has recognized the importance of a clear description of the sonographic findings in healthy children and has recently published a set of definitions ( Box 9.1 ).

BOX 9.1

Definitions for the Sonographic Components of the Normal Pediatric Joint

Definition 1

The hyaline cartilage will present as a well-defined anechoic structure (with/without bright echoes/dots) that is noncompressible. The cartilage surface can (but does not have to) be detected as a hyperechoic line.

Definition 2

With advancing maturity, the epiphyseal secondary ossification center will appear as a hyperechoic structure, with a smooth or irregular surface within the cartilage.

Definition 3

Normal joint capsule: A hyperechoic structure that can (but does not have to) be seen over bone, cartilage, and other intraarticular tissue of the joint.

Definition 4

Normal synovial membrane: Under normal circumstances, the thin synovial membrane is undetectable.

Definition 5

The ossified portion of articular bone is detected as a hyperechoic line.

Interruptions of this hyperechoic line may be detected at the growth plate and at the junction of two or more ossification centers.

For the differentiation of synovitis from physiological findings, the degree of distension of a synovial recess may also be interesting, as variable amounts of fluid can be found in the joints of healthy children as well. This has led to the publication of normative data for various joints ( Table 9.3 ). ^, It has to be noted, though, that the definition of absolute cutoffs can be difficult, and the measurement of a distension of a synovial recess is prone to errors related to joint position, transducer pressure, the precise location of measurement, and others. The published reference values should therefore serve as a guidance, and the longitudinal assessment in an individual patient may be more important than decisions based on comparison with normative data at a given point in time. These normative data, however, clearly help in the distinction between physiological and pathological findings, and the addition of Doppler findings will help in the identification of pathology.

The increased metabolic activity in the musculoskeletal system of children during growth coincides with increased blood flow into and within bones and joints. This can be detected with Doppler sonography and must be distinguished from pathological blood flow, which is an important sign of active synovitis. A significant degree of Doppler signals can be detected within the joint but not within the synovial recess. In many joints, for example, the elbow, wrist, knee, or ankle, the intracapsular space includes connective tissues, which are extrasynovial. Doppler signals within these tissues should not be interpreted as a sign of synovitis because they can be physiological ( Fig. 9.14 ). The same applies to nutrient vessels directly entering bones and thereby crossing the synovial space (feeding vessels).

Physiological Doppler flow can also be observed in the area of growth plates and epiphyseal cartilage and can be reliably imaged on sonography.

Magnetic Resonance Imaging

Physiological changes during growth, such as high-signal bone marrow changes on water-sensitive MRI sequences, irregular bone surface, and moderate amounts of joint fluid, are frequently seen at the wrist and can mimic edema, erosions, or effusions ^, ( Figs. 9.15, 9.16, and 9.17 ). Bone marrow edema-like changes have also been reported in the normal foot and pelvis. ^, It is reasonable to believe that these are growth-related changes that may affect the entire skeleton; however, to date, normal references have only been published for the temporomandibular joint, ^, the sacroiliac (SI) joints, the wrist, ^, the pelvis, and the knee joint ( Table 9.3 ).

The synovium consists of a continuous surface layer (intima) of cells and the underlying tissue (subintima). The depth of the subintimal layer can be up to 5 mm in thickness according to subtype. On MRI, all three types appear as a thin line with minimal enhancement after contrast administration. Notice that on postcontrast images, the thickness of the synovium may vary according to the timing of contrast administration. ^,

Evaluation of cartilage is of particular interest in children with juvenile idiopathic arthritis (JIA) because damage may occur early while there still is a window for repair. Three types of hyaline cartilage, articular, physeal, and epiphyseal, differ morphologically and on imaging. The articular cartilage is avascular and water-rich and contains few cells, whereas the physis is relatively cellular and vascular until around 18 months of age. As noted earlier, the epiphyseal cartilage contains vascular channels. The structural differences are mirrored on MRI, particularly on water-sensitive sequences (STIR and SE T2-weighted images) returning high signal from articular and physeal cartilage and low signal from epiphyseal cartilage ( Fig. 9.18 ). All three types return intermediate signal on T1-weighted images and intermediate to high signal on PD-weighted images ( Fig. 9.18 ).

On GRE imaging, all forms of hyaline cartilage return intermediate- or high-signal ( Fig. 9.18 ). A double-echo steady state (DESS) sequence (mixed T1/T2∗-weighted sequence), particularly when used with a fat saturation technique termed water excitation , provides excellent visualization of cartilaginous detail ( Fig. 9.19 ). With all GRE imaging, isotropic imaging allows thinner sections and, if necessary, additional multiplanar reconstructions. Ultrashort time to echo imaging (UTE) allows assessment of structures not seen on conventional T2-weighted images, such as the radial and calcified zones of the articular cartilage. Currently, UTE is the only available tool (perhaps excepting ultrasonography) for the evaluation of the osteochondral junction, which can be an important interface for nourishment of the deep articular cartilage and probably involved in the pathogenesis of osteoarthritis.

Imaging in Joint Disease

A variety of imaging features can be encountered with joint disease. Specific joint findings will depend on the child’s age at diagnosis, the duration and severity of disease, and the response to therapy. A systematic approach to interpretation of any joint imaging will ensure that the salient imaging features are considered. One popular approach is the ABCDs of joint disease, where one assesses joint a lignment, b one density and other bone changes, c artilage loss, d istribution of joint disease (whether monoarticular, oligoarticular, or polyarticular), and s oft tissue abnormalities.