Note: This copy is for your personal non-commercial use only. To order presentation-ready copies for distribution to your colleagues or clients, contact us at www.rsna.org/rsnarights.
PEDIATRIC IMAGING
1253
Imaging of Juvenile Idiopathic Arthritis: A Multimodality Approach1 SA-CME See www.rsna .org/education /search/RG
LEARNING OBJECTIVES FOR TEST 2 After completing this journal-based SACME activity, participants will be able to: ■■Describe
the clinical features and classification of juvenile idiopathic arthritis. ■■Discuss
the role of different imaging modalities in the evaluation of complex joints in patients with juvenile idiopathic arthritis. ■■Identify
the characteristic imaging findings of juvenile idiopathic arthritis.
INVITED COMMENTARY See discussion on this article by Rao (pp 1273–1275).
Elizabeth F. Sheybani, MD2 • Geetika Khanna, MD, MS • Andrew J. White, MD, MSc • Jennifer L. Demertzis, MD Juvenile idiopathic arthritis (JIA) is a heterogeneous group of diseases characterized by synovial inflammation and is the most common rheumatic complaint in children. To facilitate research and treatment, JIA has been further classified on the basis of the number of joints involved, additional symptoms, family history, and serologic findings. Imaging in patients with JIA has historically relied on radiography, which allows the accurate assessment of chronic changes of JIA, including growth disturbances, periostitis, and joint malalignment. However, radiographic findings of active inflammation are nonspecific, and, in the past, clinical evaluation has taken precedence over imaging of acute disease. Recent advances in disease-modifying therapeutic agents that can help prevent long-term disability in patients with JIA have led to greater emphasis on the detection of early joint-centered inflammation that cannot be accurately assessed radiographically and may not be evident clinically. Both contrast material–enhanced magnetic resonance (MR) imaging and Doppler ultrasonography (US) are well suited for this application and are playing an increasingly important role in diagnosis, risk stratification, treatment monitoring, and problem solving. Contrast-enhanced MR imaging is the most sensitive technique for the detection of synovitis and is the only modality that can help detect bone marrow edema, both of which indicate active inflammation. US is more sensitive than radiography for the detection of synovial proliferation and effusions and is particularly useful in the evaluation of small peripheral joints. The complexity of the temporomandibular and sacroiliac joints limits the usefulness of radiographic or US evaluation, and contrast-enhanced MR imaging is the preferred modality for evaluation of these structures. ©
RSNA, 2013 • radiographics.rsna.org
Abbreviations: ERA = enthesitis-related rheumatoid arthritis, ILAR = International League of Associations for Rheumatology, JIA = juvenile idiopathic arthritis, OMERACT = Outcome Measures in Rheumatology Clinical Trials, RAMRIS = rheumatoid arthritis MR imaging score, RF = rheumatoid factor, STIR = short inversion time inversion-recovery, TMJ = temporomandibular joint RadioGraphics 2013; 33:1253–1273 • Published online 10.1148/rg.335125178 • Content Codes: From the Mallinckrodt Institute of Radiology, Washington University School of Medicine, 510 S Kingshighway Blvd, Campus Box 8131, St Louis, MO 63110. Recipient of a Certificate of Merit award for an education exhibit at the 2011 RSNA Annual Meeting. Received August 17, 2012; revision requested November 14 and received February 8, 2013; accepted February 13. For this journal-based SA-CME activity, the authors, editor, and reviewers have no relevant relationships to disclose. Address correspondence to E.F.S. (e-mail:
[email protected]). 1
Current address: The Hospital for Sick Children, Toronto, Ontario, Canada.
2
©
RSNA, 2013
1254 September-October 2013
Introduction
Juvenile idiopathic arthritis (JIA) includes all forms of arthritis that develop before the age of 16 years, persist for at least 6 weeks, and have no identifiable cause. JIA is the most common rheumatic disease in children, with a reported prevalence of 16–150 per 100,000 children (1). JIA remains a clinical diagnosis with varied manifestations and is influenced by genetic and environmental factors. Historically, imaging evaluation for known or suspected JIA has relied primarily on radiography. However, radiographic findings such as bone erosions, joint space narrowing from cartilage destruction, and growth disturbances are irreversible findings that occur late in the course of disease. The development of improved therapeutic agents whose use can prevent joint destruction, especially when treatment is initiated early, highlights the importance of early (preradiographic) detection of inflammation. The potentially serious side effects of these newer therapeutic agents in the pediatric population underscore the importance of accurately assessing disease activity, disease progression, and treatment response. As a result, management of JIA has evolved to include greater utilization of advanced imaging techniques such as contrast material–enhanced magnetic resonance (MR) imaging and Doppler ultrasonography (US), either of which can help in (a) detecting inflammatory lesions before permanent joint destruction occurs, and (b) monitoring disease progression and treatment response to more effectively guide therapy (2). In this article, we review the most recently developed classification system for JIA and describe a rational approach for imaging of this disease entity, with emphasis on MR imaging and US. In addition, we discuss the imaging of peripheral joints in patients with JIA, as well as special considerations in the imaging of complex structures such as the cervical spine, temporomandibular joint (TMJ), and sacroiliac joints.
JIA Classification
JIA is a diagnosis of exclusion that encompasses a broad range of disease manifestations and prognoses. The most recent JIA classification system was proposed by the International League of Associations for Rheumatology (ILAR) (Table 1) in an attempt to facilitate disease management, predict outcomes, and refine research populations (3). Although portions of this classification system have been brought under scrutiny and clas-
radiographics.rsna.org
sification may continue to evolve, the radiologist should be familiar with the manifestations of the various JIA subtypes, possess a rational approach for imaging affected joints, and recognize findings related to disease prognosis to guide differential diagnosis and management in patients with JIA. Oligoarthritis, which by definition affects from one to four joints within 6 months of onset, is the most common JIA subtype and classically involves the knees or ankles of preschool-aged girls (4). Iridocyclitis is a characteristic feature of oligoarthritis that affects up to 30% of patients and necessitates routine ophthalmologic screening (1). Polyarthritis, which involves more than five joints within 6 months of onset, is further classified on the basis of serologic test results. RFpositive polyarthritis classically entails symmetric involvement of the small joints of the hands in adolescent girls. It can be thought of as early-onset, adult rheumatoid arthritis with erosive joint destruction, an uncommon finding in other JIA subtypes (Fig 1) (4). RF-negative polyarthritis may manifest at any age throughout childhood, is heterogeneous in terms of presentation and prognosis, and does not have a characteristic pattern of joint involvement. Systemic arthritis is a potentially fatal subtype of JIA and includes recurrent daily fevers in its definition. Systemic findings may include characteristic rash, lymphadenopathy, and hepatosplenomegaly. The classically symmetric and polyarticular joint involvement in systemic arthritis may not develop until late in the disease course. A subset of patients with systemic arthritis develop macrophage activation syndrome, an acute, life-threatening condition that is characterized by pancytopenia, disseminated intravascular coagulation, liver dysfunction, and sustained fever resulting in multiorgan failure. The paradoxically decreased erythrocyte sedimentation rate helps distinguish macrophage activation syndrome from a rheumatologic disease flare (1). ERA predominantly affects boys over 6 years of age and is characterized by enthesitis, most commonly of the Achilles tendon insertion, plantar fascia, and tarsus, frequently with asymmetric arthritis of the lower extremities. Unlike in other subtypes of JIA, hip involvement is common at the time of presentation in ERA (1). Spinal involvement usually does not manifest until adulthood (4). Juvenile psoriatic arthritis, like RF-negative polyarthritis, is heterogeneous in terms of presentation and prognosis (1,4). Patients with psoriatic arthritis can have large or small joint involvement and may present with dactylitis before developing typical skin changes of psoriasis (Fig 2).
RG • Volume 33 Number 5
Sheybani et al 1255
Table 1 ILAR Classification of JIA Category Systemic arthritis
Oligoarthritis Polyarthritis (RF negative) Polyarthritis (RF positive) Psoriatic arthritis ERA
Undifferentiated
Definition One or more joints affected with or preceded by at least 2 weeks of fevers that have been daily for at least 3 days, with at least one of the following: transitory rash, generalized lymphadenopathy, hepato- or splenomegaly, or serositis One to four joints affected within 6 months of onset Five or more joints affected within 6 months of onset, with a negative RF test result Five or more joints affected within 6 months of onset, with two positive RF test results at least 3 months apart within 6 months of disease onset Arthritis and psoriasis, or arthritis and two or more of the following: dactylitis, nail pitting or onycholysis, or psoriasis in a first-degree relative Arthritis and enthesitis, or arthritis or enthesitis with two or more of the following: presence or history of sacroiliac joint tenderness or inflammatory low back pain; positive HLA-B27 antigen; male over 6 years of age at onset; acute (symptomatic) anterior uveitis; or a history of ankylosing spondylitis, ERA, sacroiliitis with IBD, Reiter syndrome, or acute anterior uveitis in a first-degree relative Fits into none or at least two of the other categories
Source.—Adapted from reference 3. Note.—ERA = enthesis-related arthritis, HLA = human leukocyte antigen, IBD = inflammatory bowel disease, RF = rheumatoid factor.
Figure 1. RF-positive polyarticular JIA in an 18-year-old woman. Anteroposterior radiographs of the left (a) and right (b) hands show extensive bilateral erosive changes in the carpal and metacarpal bones (arrows) and erosions with associated periosteal reaction and soft-tissue swelling at the proximal phalanges of the left third finger, right fourth finger, and right little finger (arrowheads). There is marked joint space loss in both wrists and at several metacarpophalangeal and interphalangeal joints. This pattern of involvement is reminiscent of adult rheumatoid arthritis.
1256 September-October 2013
radiographics.rsna.org
Figure 2. Psoriatic arthritis in a 10-year-old girl who presented with dactylitis. (a) Clinical photograph demonstrates swelling and erythema of the right second toe. (Courtesy of Dana Toib, MD.) (b) Anteroposterior radiograph of the right forefoot shows diffuse soft-tissue swelling with overgrowth of the second toe, a finding of chronic inflammation that is unique to the pediatric skeleton.
Figure 3. Acute radiographic changes in an 8-year-old boy with systemic JIA and left knee pain. (a) Clinical photograph demonstrates marked soft-tissue swelling of the left knee compared with the unaffected right knee. (Courtesy of Dana Toib, MD.) (b, c) Frontal (b) and lateral (c) radiographs of the left knee demonstrate nonspecific findings of active inflammation, including soft-tissue swelling (arrows) with a suprapatellar effusion or synovial thickening (*). Note the absence of erosions or joint space loss.
RG • Volume 33 Number 5
Figure 4. RF-negative polyarticular JIA in a 10-year-old girl. Anteroposterior radiograph of the right hand demonstrates periarticular osteopenia (arrows) and soft-tissue swelling, which is most marked at the proximal interphalangeal and metacarpophalangeal joints of the index finger (arrowheads), findings that are indicative of active disease.
Imaging Evaluation
Imaging in JIA may be indicated to aid in diagnosis, evaluate the extent and severity of disease, assess treatment response, detect potential complications, and direct the administration of intraarticular therapeutic agents. Unlike the systematic protocols for monitoring adult patients with rheumatoid arthritis, there are no defined imaging protocols for JIA (5). The timing and utilization of imaging in JIA must be tailored to the individual patient, with consideration given to the strengths and weaknesses of each available modality in evaluating the joints in question and directing therapeutic decision making.
Radiography Radiography has traditionally been the mainstay of imaging in JIA. Radiographs play an important role in initial diagnosis by excluding other
Sheybani et al 1257
causes of joint pain and swelling and providing a baseline standard for follow-up. Radiographs are also useful for identifying late complications of JIA such as accelerated bone growth, premature physeal fusion, and limb length discrepancy. However, routine surveillance radiographs are not predictive of disease course and should be obtained during follow-up only when there is a change in symptoms or management (6). For rheumatoid arthritis in adults, there are well-defined radiographic scoring systems, and radiographically defined structural damage has traditionally been a key outcome in clinical trials for new therapies (5,7). The largely cartilaginous composition of the pediatric skeleton limits radiographic detection of early erosive changes in children. Preerosive signs of inflammation such as synovitis and osteitis are undetectable on radiographs. In addition, the complexities of the maturing skeleton limit standardization of radiographic scoring for JIA (8). Although multiple radiographic scoring systems for JIA have been proposed, none has been widely accepted for routine clinical use, in part due to significant inter- and intraobserver variation (9,10). Because of these limitations, clinical assessment of joint function and disability takes precedence over radiographic findings. Radiographic findings in early-stage JIA are often nonspecific. The most commonly encountered radiographic findings include soft-tissue swelling, joint effusion, and osteopenia (Fig 3). Osteopenia tends to be periarticular in the early stages of disease secondary to joint inflammation (Fig 4), whereas in advanced-stage JIA it can be diffuse due to decreased physical activity or steroid administration (9,11). Persistent synovial inflammation can result in erosions, although these are less common in JIA than in adult rheumatoid arthritis because children have a large amount of epiphyseal cartilage, and significant cartilage destruction must occur before the bone is affected and erosions become radiographically apparent. In addition, the vascularization of immature cartilage provides some reparative capacity not found in adults.
1258 September-October 2013
radiographics.rsna.org
Figure 5. Disease progression in a 12-year-old girl with a 7-year history of systemic JIA. (a) Anteroposterior radiograph of the right hip shows uniform joint space narrowing with small erosions of the femoral head. (b) Radiograph obtained 2 years later shows severe joint space loss (arrow) due to progressive, irreversible cartilage destruction. Acetabular and femoral head erosions (arrowhead) have also increased.
The prevalence of erosions and joint space loss has been shown to be higher in children with polyarticular JIA than in those with oligoarticular JIA (12). Although erosions can occur anywhere along the articular surface, they are most common at the sites of synovial reflection and ligament insertion, where there is less overlying protective cartilage. Joint space narrowing occurs due to thinning and gradual loss of the articular cartilage and is an irreversible finding (Fig 5). For evaluation of the lower extremities, weight-bearing radiographs are recommended over supine images for detecting joint space loss, limb length discrepancy, or joint malalignment, a particularly important consideration given that the knee is the most commonly involved joint in JIA. Periostitis is more common in JIA than in adult rheumatoid arthritis (6). It most commonly occurs along the metacarpal and metatarsal bones, giving the bones an enlarged, squaredoff appearance (6,13). Periostitis should not be confused with intraarticular, capsular, or periarticular calcifications, which can occur following therapeutic steroid injection (Fig 6). The cause of these periarticular calcifications is unknown. They most commonly occur around the knee, are composed of hydroxyapatite, and may be related to trauma and tissue necrosis (14,15). Radiographic findings of advanced JIA include ankylosis, growth disturbances, and joint malalignment. Ankylosis, a feature of severe, long-standing disease in adults, has been reported to occur
Figure 6. Iatrogenic periarticular calcification in a 9-year-old girl with a 2-year history of oligoarthritis who was treated with multiple intraarticular injections of triamcinolone. Anteroposterior radiograph of the proximal interphalangeal joint of the right fourth finger shows extraarticular calcifications along the ulnar aspect of the joint. Joint-centered soft-tissue swelling is also present.
within 3–5 years of disease onset in JIA. It is most commonly seen in the carpal and tarsal bones and the apophyseal joints of the upper cervical spine (Fig 7) (16,17). Growth disturbances are distinct radiographic findings of JIA as opposed to adult
RG • Volume 33 Number 5
Sheybani et al 1259
Figure 7. Chronic radiographic changes in a 21-year-old woman in whom a diagnosis of polyarticular JIA had been made at 2 years of age. Posteroanterior radiograph of the right wrist shows erosions (arrows) at the radial articular surface and metacarpal bases and along the carpal bones, resulting in crenated margins. There is diffuse joint space loss with ankylosis (arrowheads) of the second metacarpal, trapezium, trapezoid, capitate, and scaphoid bones. Note the periarticular osteopenia and severe negative ulnar variance caused by premature closure of the ulnar physis.
Figure 8. Accelerated bone maturation in a 4-year-old girl with a 1-year history of oligoarticular JIA. Posteroanterior radiographs of the left (a) and right (b) wrists show asymmetric, advanced maturation and overgrowth of the carpal bones and distal radial epiphysis on the right side; the left side is unaffected. Note that the trapezoid ossification center is present on the right side (arrow) but has not yet appeared on the left side.
arthritis (13). These changes reflect unique features of the immature skeleton, including growth potential, high cartilaginous content, and epiphyseal vascularization. Growth disturbances are more likely in patients who are young at the time of disease onset. Localized hyperemia may result in epiphyseal enlargement, accelerated maturation, and osseous overgrowth. Around the knee joint, epiphyseal enlargement causes widening of the intercondylar notch with squaring of the lower pole of the patella, findings that are classic for JIA but are also seen with hemophiliac arthropathy. During the early stages of disease, the accelerated growth results in a relative increase in the length of the affected
limb. In the hands, premature ossification of the carpal bones can result in a bone age greater than the chronologic age (Fig 8). In longstanding disease, accelerated maturation results in premature physeal fusion and epiphyseal destruction with resultant limb shortening or joint malalignment. Joint malalignments such as subluxation, dislocation, and flexion and extension deformities can occur in any joint but are most commonly seen in the hands and feet. In contradistinction to the ulnar deviation seen at the wrist in adult rheumatoid arthritis, radial deviation at the wrist is typically seen in JIA (6).
1260 September-October 2013
radiographics.rsna.org
Figure 9. Radiographically occult active inflammation in an 8-year-old girl with RF-negative polyarthritis. (a) Anteroposterior radiograph of the left wrist shows normal findings, with the exception of an incidental lunotriquetral coalition. (b) Coronal contrast-enhanced fat-suppressed T1-weighted MR image demonstrates diffuse carpal bone marrow edema (*), a finding that is consistent with osteitis, as well as enhancing synovitis of the distal radioulnar (arrow) and intercarpal joints.
Despite the role of radiography as the standard of reference for joint evaluation in JIA, radiographic detection of inflammatory preerosive changes is limited (Fig 9) (9,11). Long-standing synovitis may lead to osteocartilaginous damage and permanent functional impairment, affecting up to one-third of JIA patients into adulthood (11). Early detection of synovial inflammation, cartilage damage, and erosive changes is essential to initiating disease-modifying therapy and preventing long-term disability. MR imaging and US are uniquely suited to this goal. Both modalities are superior to radiography in the evaluation of synovial and cartilaginous changes and are increasingly being used in disease classification, prognosis, and treatment monitoring.
MR Imaging With its multiplanar capability and excellent bone and soft-tissue contrast resolution, MR imaging is well-suited for imaging patients with JIA. MR imaging allows comprehensive evaluation of the synovium, articular cartilage, growth cartilage, bone marrow, cortical bone, and soft tissues. Contrast-enhanced MR imaging is the most sensitive imaging technique for detecting synovitis (9). In early JIA, MR imaging can help detect synovitis before it is apparent at physical examination, a potentially important prognostic indicator (18,19). MR imaging is the only imaging modality that can demonstrate bone marrow edema (6,9),
and, although its prognostic significance in JIA has not been clearly defined, in adult rheumatoid arthritis bone marrow edema is a predictor of future erosions (20). Therefore, the presence of bone marrow edema in the setting of JIA is considered a preerosive abnormality and an indication for initiating therapy to prevent permanent joint damage (11). MR imaging has been shown to help detect more than twice as many bone erosions in the wrist as either US or radiography (Fig 10) (21) and more than twice as many cases of sacroiliitis as radiography in a pediatric population (22). It can also help identify radiographically occult extraarticular inflammatory lesions such as tenosynovitis and enthesitis (9). MR imaging is nonirradiating, an important consideration in this radiosensitive population, although when a joint is being evaluated with MR imaging, correlative radiographs should be obtained to document findings and for future reference during treatment monitoring. As with radiography, there are no fully validated systems for monitoring JIA with MR imaging (11). In adults, the Outcome Measures in Rheumatology Clinical Trials (OMERACT) study group has developed a semiquantitative scoring system and imaging protocol for the assessment of synovitis, bone marrow edema, and bone erosions in the wrist and metacarpophalangeal joints of rheumatoid arthritis patients (rheumatoid arthritis MR imaging score [RAMRIS]) (5). RAMRIS has been shown to have acceptable applicability to children despite differences in the pediatric and adult skeletons, and the three
RG • Volume 33 Number 5
Sheybani et al 1261
Figure 10. Improved detection of erosions with MR imaging of the wrist in a 16-year-old girl with polyarticular JIA. (a) Posteroanterior radiograph through the left wrist demonstrates diffuse osteopenia with a well-defined erosion at the base of the hamate bone (arrow). Additional erosions are present but are difficult to delineate. (b) Coronal T1-weighted MR image demonstrates the erosion of the hamate bone (arrow) and reveals erosions in the scaphoid bone and the base of the second metacarpal bone (arrowheads). Additional images demonstrated erosions in nearly all of the carpal bones. (c) Sagittal contrast-enhanced fatsuppressed T1-weighted MR image shows uniform enhancement of the erosions involving the capitate bone and the base of the third metacarpal bone (arrowheads), and tenosynovitis in the extensor tendons (arrow), findings that are consistent with active inflammation.
principal scored findings are often extrapolated to joints other than the wrist and hand (23). MR imaging should be performed on a highfield-strength magnet with a local coil closely matched in size to the affected area to achieve an acceptable signal-to-noise ratio. Spin-echo or fast spin-echo (T1-weighted) MR images (short repetition time, short echo time) are used to assess bone marrow and erosions. Water-sensitive fat-suppressed (fast spin-echo or short inversion time inversion-recovery [STIR] T2-weighted) MR images are used to evaluate joint and tenosynovial fluid, cartilage, marrow
edema, and tendons. The administration of gadolinium-based contrast material is essential for distinguishing active synovial inflammation from nonenhancing joint effusions or fibrotic (noninflamed) pannus; normal synovium demonstrates minimal to no enhancement. Contrastenhanced fat-suppressed T1-weighted images (short repetition time, short echo time) should be obtained within 5–10 minutes of injection. Beyond this time, diffusion of contrast material into the joint limits differentiation between enhancing synovium and adjacent joint fluid (24). A combination of water-sensitive and postcontrast sequences (short repetition time, short echo time) performed in at least two orthogonal planes is essential for complete evaluation of any joint in the setting of JIA. This may be achieved
1262 September-October 2013
by either imaging in perpendicular planes with two-dimensional sequences or performing a three-dimensional sequence with isometric voxels that can be reconstructed in other planes. Assessment of articular cartilage with MR imaging has traditionally relied on a morphologic description of detected abnormalities. A relatively recent innovation in the evaluation of JIA is T2 relaxation time cartilage mapping, which reflects biochemical changes in water and collagen content and in tissue anisotropy in the extracellular matrix of articular cartilage. T2 mapping of articular cartilage in the knee has been shown to help detect microstructural changes caused by inflammation before morphologic changes can be detected at conventional MR imaging (25). Kim et al (26) reported a progressive increase in T2 relaxation time for distal femoral articular cartilage during follow-up of children with recently diagnosed JIA, despite improving clinical assessment. More recently, T2 relaxation mapping techniques have been successfully applied in studying the cartilage of the small joints of the hands (27). Despite these benefits, MR imaging examinations are lengthy, require the intravenous administration of contrast material to maintain sensitivity, and frequently entail sedation of pediatric patients to obtain a diagnostic study. Moreover, imaging is restricted to a single joint or area of the body to achieve adequate contrast and spatial resolution. Because of these limitations, MR imaging is presently reserved for critical therapeutic decision making and problem solving in unusual cases, with the exception of the evaluation of complex structures such as the TMJ and sacroiliac joints, for which MR imaging is the current standard of reference (11).
Ultrasonography US is a versatile modality for imaging patients with JIA. It is nonionizing, relatively inexpensive, and readily accessible; allows a survey of multiple joints; and does not require sedation. Although US is operator dependent, in experienced hands it provides reliable assessment of synovial proliferation, joint effusion, cartilage thickness, cartilage and cortical erosions, and tenosynovitis. US allows direct visualization of the articular cartilage, which is normally seen as a hypoechoic structure with a smooth outline over the bone surface. US can also be used to facilitate joint aspiration or therapeutic injection. There are no standardized US protocols for evaluating patients with JIA, but abnormalities such as erosions, tenosynovitis, and enthesitis must be documented in two perpendicular
radiographics.rsna.org
planes. Therefore, orthogonal imaging of each area of concern is recommended. US evaluation of joints should be performed with highfrequency (12–15-MHz) linear transducers that provide sufficient resolution for evaluation of articular cartilage and intraarticular diseases such as erosions (6). Incorporation of color or power Doppler imaging provides diagnostic information regarding synovial vascularity and hyperemia (6,28). The availability of US contrast agents may further improve detection and assessment of synovial proliferation and hyperemia (29). Compared with radiography, US is more sensitive for detecting joint effusions and synovial thickening and has been shown to be particularly helpful in evaluating the small joints of the hands and feet (30,31). Several studies have also demonstrated that US helps accurately detect subclinical synovitis and, in conjunction with physical examination, is useful in evaluating response to intraarticular therapy (32–35). Timely detection of subclinical synovitis is particularly important in JIA, not only because early detection and treatment may prevent long-term disability, but also because the ILAR subtypes are classified on the basis of the number of joints involved within 6 months of disease onset (3). Additionally, identification of subclinical joint involvement at US may allow patient inclusion in clinical trials for newer biologic agents in which polyarticular disease and active inflammation are required for enrollment (13,36). It has been suggested that, given the low sensitivity of physical examination for active synovitis, asymptomatic joints should be screened with US in patients with at least one actively inflamed joint. However, the prognostic implications of USidentified synovitis in the absence of clinically active disease are not well established. In one study in which patients were followed up for 6 months, only 35.7% of patients with US-identified synovitis went on to develop clinically apparent disease. In the remaining patients, mild hyperemia was the only abnormal US finding, which suggests that mild hyperemia alone is not sufficient to diagnose active synovitis (37). US evaluation does have its disadvantages. Operator variability and the scarcity of data regarding standardized imaging protocols limit the reproducibility, interpretation, and comparison of studies. In large joints, visualization of the entire articular surface may be limited by adjacent shadowing bone, and sound beam attenuation limits the amount of anatomic detail seen at imaging of deeper structures. US cannot be used to evaluate complex joints such as the sacroiliac joints and the TMJ, and it has high false-negative rates in the detection of subtalar disease (3,37).
RG • Volume 33 Number 5
Sheybani et al 1263
Table 2 US and MR Imaging Definitions for Joint and Soft-Tissue Disease in Inflammatory Arthritis Abnormality Synovitis
MR Imaging Definition
US Definition
Increased synovial enhancement with synovial thickness greater than the width of normal synovium Nonenhancing T1-hypointense and T2-hyperintense intraarticular fluid
No standardized definition; see “Synovial hypertrophy” Abnormal, generally hypoechoic displaceable or compressible intraarticular material that exhibits no Doppler signal
Synovial hypertrophy
No standardized definition; see “Synovitis”
Tenosynovitis
Increased water content or abnormal enhancement of a tendon sheath
Periarticular inflammation
Increased water content or abnormal enhancement at extraarticular sites including periosteum (periostitis) and entheses (enthesitis), but not tendon sheaths (tenosynovitis) See “Periarticular inflammation”
Abnormal, generally hypoechoic nondisplaceable and poorly compressible intraarticular tissue that may exhibit Doppler signal Hypo- or anechoic thickened tissue with or without fluid within the tendon sheath that is visible in two perpendicular planes and may exhibit Doppler signal See “Enthesopathy”
Synovial fluid
Enthesopathy
Bone edema
Bone erosion
Ill-defined lesion in trabecular bone with signal characteristics consistent with increased water content (low T1 and high T2 signal) Sharply marginated, juxtaarticular, T1-hypointense and T2-hyperintense bone lesion that disrupts the cortical surface in at least one plane
Abnormally hypoechoic or thickened tendon or ligament at its bone attachment that is visible in two perpendicular planes and may exhibit Doppler signal or bone changes Not seen at US
Intraarticular discontinuity of the bone surface that is visible in two perpendicular planes
Source.—Adapted from references 38 and 39.
Joint-specific Findings and Recommendations Peripheral Joints Although the patterns of peripheral joint involvement vary with JIA subtype, the imaging findings in active and chronic disease are similar throughout the large and small joints of the appendicular skeleton. In the setting of JIA, the primary goal of imaging is to detect active inflammation that may be amenable to the administration of either site-specific or systemic therapy to prevent irreversible joint damage. Generally, both US and MR imaging will help make this determination in the large and small joints of the extremities. With either modality, important findings to document include the presence or absence of synovitis and synovial hypertrophy, joint effusion, tenosynovitis, periarticular inflammation such as enthesitis and periostitis, and bone erosions. Bone marrow edema
can be detected with MR imaging, whereas US is not helpful in this regard. Because there are no standardized, validated definitions for these findings in the pediatric population, the definitions for these findings in adults that have been proposed by the OMERACT study group for both US and MR imaging are extrapolated to the pediatric population (Table 2) (11,23). At MR imaging, synovitis is defined as thickened synovium with avid enhancement on early postcontrast images (Fig 11) (40). Synovitis cannot be reliably distinguished from adjacent joint fluid on nonenhanced images; both entities are hypointense with T1-weighted sequences and hyperintense with water-sensitive sequences. Although synovitis in isolation is a nonspecific finding, in the setting of inflammatory arthritis it represents active joint inflammation even in patients who are in clinical remission (5,41,42).
1264 September-October 2013
radiographics.rsna.org
Figure 11. Synovitis in a 17-year-old girl with RF-negative polyarticular JIA. (a) Coronal fatsuppressed T2-weighted MR image demonstrates hyperintense synovial fluid (*) and thickening (arrow), with bone marrow edema (arrowhead) in the lateral tibial plateau. (b) Coronal contrastenhanced fat-suppressed T1-weighted MR image more clearly depicts nodular enhancing synovitis (arrow). Contrast enhancement is essential for differentiating between nonenhancing joint effusion and enhancing synovitis.
Figure 12. Rice bodies in a 3-year-old girl with oligoarticular JIA. (a) Sagittal fat-suppressed T2-weighted MR image of the knee demonstrates a large joint effusion with hypointense debris layering dependently within the effusion. (b) Contrast-enhanced fat-suppressed T1-weighted MR image shows enhancing synovitis but no enhancement of the layering debris; the latter finding is consistent with rice bodies secondary to chronic inflammation.
RG • Volume 33 Number 5
Sheybani et al 1265
Figure 13. Synovial hypertrophy in a 20-year-old woman in whom polyarticular JIA was diagnosed in childhood. (a) Longitudinal gray-scale US image through the dorsal aspect of the wrist demonstrates thickened, hypoechoic, noncompressible intraarticular tissue consistent with synovial hypertrophy (*) deep to the normal-appearing extensor tendon (arrowheads). (b) Color Doppler US image through the same region demonstrates marked synovial hypervascularity, a finding that is indicative of inflammation.
Figure 14. Tenosynovitis. (a) Sagittal fat-suppressed T2-weighted MR image of the third metacarpal bone and third finger of the right hand in a 5-year-old girl with oligoarticular JIA demonstrates fluid signal intensity within the flexor tendon sheath (arrows), a finding that is consistent with tenosynovitis. There is also edema in the subcutaneous fat along the volar aspect of the hand and finger. (b) Longitudinal color Doppler US image of the flexor tendon of the right third finger in a different patient demonstrates tenosynovitis, with hyperemic, hypoechoic noncompressible tissue within the tendon sheath, and tendinopathy (*) with a thickened tendon and loss of normal fibrillar echotexture.
With progressive inflammation, the synovium can proliferate to form pannus that can be seen on precontrast T2-weighted images as hypointense frondlike material outlined by high-signal-intensity joint fluid. Small fragments of the hypertrophic synovium may undergo fibrinoid necrosis and slough into the joint, resulting in the formation of “rice bodies” (Fig 12), a term that reflects the macroscopic resemblance of these fragments to grains of polished rice. Rice bodies are indicators of chronic synovial inflammation when seen in JIA and can also be found in chronic low-grade infections such as tuberculous arthritis and nonpigmented villonodular synovitis. Synovial hypertrophy at US is defined by poorly compressible intraarticular tissue that is most frequently hypoechoic relative to subcutaneous fat and may demonstrate Doppler signal (Fig 13) (38). Although the presence of Doppler signal is not required for the US diagnosis of synovitis, the significance of isolated synovial hypertrophy with-
out Doppler signal has not been validated, and the disappearance of Doppler signal may indicate resolving or inactive disease (11). At MR imaging, periarticular inflammation (which includes both periostitis and enthesitis) and tenosynovitis demonstrate signal intensity consistent with increased water content or increased contrast enhancement. Unlike with synovitis, however, the presence of abnormally increased water content alone (high signal intensity on water-sensitive T2-weighted and STIR images, low signal intensity on unenhanced T1-weighted images) is sufficient for the diagnosis of tenosynovitis (Fig 14) (39). Although there are only limited data on the MR imaging findings of tenosynovitis in children, the literature on adult patients suggests that postcontrast images are crucial to evaluation for early tenosynovitis, since fluid-sensitive sequences alone have fairly low sensitivity in the
1266 September-October 2013
radiographics.rsna.org
Figure 15. Tenosynovitis in a 6-year-old girl with psoriatic arthritis. (a) Clinical photograph of the left ankle demonstrates softtissue swelling, a finding that was thought to be secondary to an effusion. (Courtesy of Dana Toib, MD.) (b) Anteroposterior radiograph demonstrates medial soft-tissue swelling of the ankle. (c) Coronal contrastenhanced fat-suppressed T1-weighted MR image reveals that the soft-tissue swelling is secondary to marked tenosynovitis (arrows) of the posterior tibialis, flexor digitorum longus, and flexor hallucis longus tendons and periarticular inflammation (*) with synovitis of the ankle and subtalar joints (arrowheads).
acute and subacute stages of disease (43). At US, tenosynovitis manifests with hypoechoic or anechoic thickening of the tendon sheath (Fig 14) (38). Doppler signal may be present but is not required for diagnosis, and fluid within the tendon sheath is a variable finding. Tenosynovitis is most commonly seen around the ankle joint and along the extensor tendons of the wrist. Tenosynovitis of the ankle tendons is often clinically occult or may
mimic an ankle joint effusion at physical examination (Fig 15) (44). The term enthesitis refers to inflammation of tendons or ligaments at the site of bone attachment. It is most commonly seen at the insertion of the Achilles tendon on the posterior calcaneus. MR imaging findings may include soft-tissue edema, increased T2 signal or thickening of the tendon or ligament at its insertion, subjacent bone marrow edema, and distention of adjacent bursae. At US, enthesitis is characterized by hypoechoic
RG • Volume 33 Number 5
Sheybani et al 1267
Figure 16. Erosions and edema in the elbow of a 15-year-old boy with polyarticular JIA. (a) Sagittal contrast-enhanced fat-suppressed T1-weighted MR image through the right elbow demonstrates severe synovitis (arrows) with erosions and osteitis. The margins of the capitellar erosion (arrowhead) are well defined and disrupt the articular surface, whereas trochlear osteitis (*) is more ill defined without cortical destruction. (b) Lateral radiograph demonstrates erosions of the radial head and capitellum (arrowheads), uniform radiocapitellar joint space narrowing, and displacement of the anterior and posterior fat pads (arrows) owing to proliferative synovitis.
Figure 17. Epiphyseal cartilage hyperemia in a 4-year-old girl with oligoarticular JIA. Sagittal contrast-enhanced fat-suppressed T1-weighted MR image through the right knee demonstrates cartilage hyperemia in the distal femoral epiphysis with a spoke-wheel pattern of vascular enhancement (arrows). Synovitis and a joint effusion are also present.
thickening or loss of the normal fibrillar structure of the tendon or ligament at its insertion. Additional US features include calcifications and bone changes (eg, enthesophytes, erosions, or cortical ir-
regularity). Hyperemia seen at Doppler evaluation may be associated with these findings (38). Bone marrow edema is visible only at MR imaging. It is an ill-defined enhancing lesion in trabecular bone that demonstrates increased signal intensity with water-sensitive sequences and decreased signal intensity with T1-weighted sequences. Histologically, the term edema is a misnomer. The MR imaging findings reflect true osteitis with an inflammatory cellular infiltrate in bone, rather than increased water content (5). Like bone marrow edema, erosions have low T1 and high T2 signal at MR imaging; however, erosions demonstrate well-defined margins (Fig 16). At both US and MR imaging, erosions must be juxtaarticular, visible in two planes, and demonstrate cortical disruption in at least one plane. Although erosions may be a chronic finding, enhancement following contrast material administration suggests the presence of active, hypervascularized pannus within the bone defect (5). MR imaging is more sensitive than radiography for the detection of erosions. Unresolved joint inflammation can result in irreversible cartilage degradation and destruction. In the skeletally immature child, a “spoke-wheel” pattern of enhancement may be seen in growth cartilage at MR imaging secondary to hyperemia caused by long-standing inflammation (Fig 17), resulting in epiphyseal overgrowth. It is unusual to see cartilage erosions in young children due to thick, reparative cartilage with a robust blood
1268 September-October 2013
radiographics.rsna.org
supply, although with increasing age and persistent or progressive inflammation, cartilage thinning and erosions may occur.
Cervical Spine The cervical spine shows characteristic changes in children with JIA. The prevalence of clinical findings of cervical inflammation in affected patients has been reported to be approximately 60% (45). The cervical spine is more commonly involved in children with polyarticular or systemic JIA than in those with oligoarticular JIA (45,46). Clinical symptoms of cervical spine involvement include neck stiffness and limited range of motion, especially upon extension and lateral flexion. The characteristic radiographic finding of JIA in the cervical spine is ankylosis of the apophyseal joints (47). The ankylosis usually involves C2–3, although it often involves multiple levels. The vertebral bodies are commonly hypoplastic in both anteroposterior and transverse dimensions (“juvenile cervical vertebrae”) with concomitant narrowing of the intervertebral disk space, typically seen at the same level as the apophyseal joint ankylosis (Fig 18). Odontoid process erosions due to atlantoaxial synovial proliferation can be seen in polyarticular JIA. Inflammation at the atlantoaxial articulation with subsequent ligamentous insufficiency can result in widening of the atlantodental interval (the normal interval in children is 5 mm or less) and atlantoaxial instability, which is best evaluated on flexion and extension radiographs. Patients with polyarticular JIA are also at risk for atlantoaxial impaction, which is generally seen in young adulthood. Atlantoaxial impaction is diagnosed if the tip of the dens lies more than 4.5 mm above the McGregor line (the “line” connecting the posterior margin of the hard palate to the undersurface of the occipital bone) (48–50).
Temporomandibular Joints The TMJ can be involved in any of the JIA subtypes and is affected in 17%–87% of this patient population (51–53). Given the morphogenesis and complex anatomy of the TMJ, this joint is at especially high risk for growth disturbances when affected by inflammatory arthritis, and long-term involvement may result in both poor aesthetic and functional outcomes (54–56). It is well documented that clinical symptoms, such as pain, may not be present even in the setting of severe erosive TMJ disease, and subjective symp-
Figure 18. Cervical spine changes in a 14-year-old patient with long-standing polyarticular JIA. Lateral radiograph demonstrates ankylosis of the posterior elements of the cervical spine (*). The small, narrow vertebral bodies and disk space narrowing seen at the levels affected by ankylosis are characteristic findings in JIA, as is antegonial notching of the mandible (upward curving of the inferior surface of the mandible anterior to the angular process) (arrowhead).
toms may lead to underestimation of the degree of early inflammation (57–59). In one study, a pediatric rheumatologist who examined affected patients for signs of synovitis, including swelling, tenderness, and limited range of motion, was able to accurately diagnose active TMJ inflammation in only 58% of cases. Orthodontic examination, including a detailed questionnaire, measurements of mouth opening and deviation, and evaluation for crepitus and pain, allowed identification of only 39% of patients with active inflammation without bone deformity (60). As with other joints, radiography can demonstrate chronic erosions of the TMJ but is unable to help identify active synovial inflammation, osteitis, or joint effusion (Fig 19) (61). Although technically appealing, US is not recommended for diagnosis or surveillance of TMJ inflammation.
RG • Volume 33 Number 5
Sheybani et al 1269
Figure 19. Limitations of radiographic evaluation of the TMJs in a 12-year-old girl with oligoarticular JIA. (a) Openmouth radiographs show flattening and irregularity of the left mandibular condyle (arrow). No erosions are seen on the contralateral side. (b) Coronal contrast-enhanced fat-suppressed T1-weighted MR image helps confirm flattening of the left mandibular condyle. Actively inflamed synovitis (arrowhead) and osteitis (black arrow) are also seen. Note the erosions of the right mandibular condyle (white arrows), which were not evident radiographically, and the presence of synovitis.
guided injection of the TMJ, although the use of MR imaging– and US-guided techniques has also been reported (63,64).
Sacroiliac Joints
Müller et al (60) demonstrated that although US is the most specific imaging modality for confirming TMJ involvement, its level of sensitivity (33%) is unacceptable. US is able to help detect only the late changes of TMJ arthritis, such as destruction and disk dislocation, and is therefore not recommended for TMJ surveillance in patients with JIA. Contrast-enhanced MR imaging is the standard of reference for evaluation of the TMJ. Certain patients, such as those who present at a very young age (
