SEARCH

SEARCH BY CITATION

Introduction to the use of in-office magnetic resonance imaging (MRI)

  1. Top of page
  2. Introduction to the use of in-office magnetic resonance imaging (MRI)
  3. Basic MRI concepts
  4. Safety considerations
  5. Available in-office extremity MRI systems
  6. Reimbursement and obstacles to payment
  7. Predictive value of high-field MRI
  8. Review of published data regarding comparison of low- and high-field units and radiographic correlation
  9. Management considerations
  10. Summary
  11. Future research directions
  12. REFERENCES

Research on rheumatoid arthritis (RA) patient cohorts has demonstrated that disability is associated with joint pain and inflammation in the early years of disease, but, in the long term, disability and inability to function in activities of daily living correlate closely with structural damage (1). Over the last decade, new therapies that effectively prevent such joint cartilage and bone loss have become available. Clinical trials using validated measures of radiographic progression, such as the total Sharp score (2), have demonstrated the ability of methotrexate, newer disease-modifying antirheumatic drugs, and biologic agents inhibiting tumor necrosis factor and interleukin-1 to slow or prevent joint damage (3–5). Based on the results of these trials, many rheumatologists began to utilize radiography more frequently to stage patients at the onset of disease as well as to monitor response to therapy.

Although MRI technology has been available for the last 30 years, experience in patients with RA is limited (6–8). Over the last decade, a number of reports have described the efficacy of MRI in patients with RA, demonstrating the ability to document erosions and synovitis with a greater sensitivity than conventional radiography (9–12). These studies have shown that bone and soft tissue abnormalities that are seen on MRI often progress to radiographic erosive disease. Such investigations were generally conducted using high-field strength magnets (1.5T) on small numbers of patients, and often provided little or no clinical correlation.

Recently, extremity MRI units, both low-field (0.2T) and high-field (1.0T), have become commercially available, and studies of the use of these machines have been presented as abstracts or published in peer-reviewed journals. These machines are small enough for use in the clinic and are being purchased by rheumatologists for use in the diagnosis and management of RA. Since these machines are expensive and the cost to the patient and the practice may be significant, the American College of Rheumatology (ACR) established a task force to review the evidence to date evaluating the benefits and limitations of these extremity units. This task force consisted of academic and private-practice rheumatologists along with experienced musculoskeletal radiologists. The task force identified the following questions that needed to be addressed in this review, and this document will summarize our findings:

  • 1
    What extremity MRI systems are presently commercially available and what is the cost of these machines? What are the ranges of reimbursement and what, if any, obstacles exist for payment?
  • 2
    How do the images obtained with the low-field extremity MRI systems compare with those obtained with the high-field systems? How do these systems compare in quality, reproducibility, operator dependence, and cost?
  • 3
    Considering that intravenous MRI contrast agents are not being used routinely in the clinic setting with low-field extremity MRI, how does this impact the accuracy and precision of these studies?
  • 4
    What is the predictive value of MRI abnormalities, including synovitis, bone marrow edema, and erosion, for the subsequent development of radiographic erosion?
  • 5
    What data are presently available to support change in disease management based on MRI findings over and above other clinical, laboratory, and radiographic findings?
  • 6
    What data are available to indicate that abnormalities revealed by MRI are predictive of poorer clinical outcomes or long-term disability?

Titles and abstracts of all articles identified were manually reviewed to identify all those that addressed the use of MRI in the evaluation of RA of the hands or feet. The reference lists of all selected reports were then reviewed to identify additional relevant articles. It was decided to only include data from articles published in peer-reviewed journals, and therefore, papers published only as abstracts were not included in the analysis. The data retrieval concluded on September 1, 2005, and the information presented here is subject to change as further information becomes available. A glossary of relevant terms is provided in Table 1.

Table 1. Glossary of relevant magnetic resonance imaging (MRI) terms*
  • *

    Information provided by Aaron D. Sodickson, MD, PhD (Brigham and Women's Hospital, Boston, MA).

CryogenLow-temperature liquid helium or nitrogen used to cool superconducting magnets.
Field strengthStatic magnetic field within the scanner, measured in Tesla.
Field of view (FOV)The distance of anatomic coverage in a given imaging direction, determined in part by the size of the radiofrequency coil being used.
Fringe field“Stray” magnetic field extending outside the imaging bore of the magnet. The distance this field extends outside the bore is a major safety consideration in designing the size and shielding requirements of MRI rooms.
GradientVariation in magnetic field strength with change in distance, used to determine voxel location when making an image. Measured in milli-Tesla per meter (mT/m).
Image planeMay be arbitrarily selected; most common planes include axial, coronal, and sagittal, although intermediate oblique planes may be utilized.
MatrixThe number of “in-plane” pixels along each given image direction. In combination with FOV, determines the in-plane image resolution.
Pulse sequencesTiming of MRI parameters (radiofrequency pulse strength and spacing, magnetic field gradients, and signal collection) used to create MR images with varying degrees of tissue contrast.
Radiofrequency (RF)Energy deposited in the patient in order to produce MRI signals (usually in the megahertz frequency range at typical magnetic field strengths used). A side effect is unwanted heating of tissues, which limits the amount of allowable energy deposition.
Selective fat saturationAlso known as chemical shift fat saturation, a method of removing fat signal based on the different signal frequencies of fat and water.
Slice thicknessThe through-plane voxel dimension.
Spatial resolutionDefinition of the smallest structures that can be differentiated on an image, generally related to pixel or voxel dimensions, although voxels can be interpolated to artificially increase display resolution from the true image resolution. True in-plane resolution equals field of view divided by matrix.
STIR”Short tau inversion recovery” pulse sequence; a popular and robust method used for suppression of MRI signal from fat.
Tesla (T)Unit of magnetic field strength. 1 Tesla equals 10,000 gauss (the earth's magnetic field strength is ∼0.5 gauss).
Voxel“Volume element,” the 3-dimensional size of each point in an image, generally determined by 2 in-plane pixel dimensions (in turn determined by FOV and matrix) and the slice thickness.

Basic MRI concepts

  1. Top of page
  2. Introduction to the use of in-office magnetic resonance imaging (MRI)
  3. Basic MRI concepts
  4. Safety considerations
  5. Available in-office extremity MRI systems
  6. Reimbursement and obstacles to payment
  7. Predictive value of high-field MRI
  8. Review of published data regarding comparison of low- and high-field units and radiographic correlation
  9. Management considerations
  10. Summary
  11. Future research directions
  12. REFERENCES

Magnet system configuration and field strength.

MR imaging systems are available in several configurations and field strengths. The magnet configuration refers to the type and design of the magnet and may be whole-body cylindrical, whole-body open, or extremity-only. Magnet field strength may be high-field (≥1.0T), mid-field (0.5T–1.0T), or low-field (<0.5T). High-field magnets are superconducting magnets that require cryogens (e.g., liquid helium). Low-field magnets are most commonly permanent magnets or electromagnets that do not require cryogens. Mid-field magnet systems are available in superconducting or non-superconducting configurations, depending on the manufacturer.

Whole-body cylindrical magnet systems are the most common. These magnets provide a homogeneous magnetic field, typically over a sufficiently large field of view, which allows all joints and body parts to be imaged. Most whole-body cylindrical magnets are part of high-field MRI systems of ≥1.0T field strength. The disadvantages of cylindrical magnets are limitations in patient positioning and the requirement that much of the patient must enter the magnet, which may be difficult for patients who are claustrophobic.

Whole-body open magnet systems are configured with a noncylindrical, commonly C-shaped geometry. These systems offer greater ease of patient positioning and a lower likelihood of inducing claustrophobia. The disadvantage is that most of these systems are low-field or mid-field magnets with limited image signal-to-noise ratio (SNR). These SNR considerations are described below.

Extremity-only magnets are designed to image the peripheral joints of the upper and lower extremities. Both high- and low-field extremity MRI systems are currently available. The low-field systems have a U-shaped, or slot-opening geometry, whereas the high-field extremity system utilizes a cylindrical magnet. The advantages of extremity systems include patient comfort, easy site preparation, lower up-front and operating costs, and minimal or no cryogens. Only the extremity being imaged is placed within the magnet. As a result, the risk of patient claustrophobia is virtually eliminated. The extremity magnets have smaller fringe magnetic fields and thereby require less space and magnetic shielding than whole-body magnets. The disadvantages of extremity magnets include the limited size of the homogeneous magnetic field, limited choices in image acquisition techniques, and, for low-magnetic-field systems, no ability to obtain frequency-selective fat-saturation images. The lower SNR of low-field systems may also limit the spatial resolution that can be obtained when compared with high-field systems. The relatively inhomogeneous magnetic field of these smaller magnets restricts the field of view of the images. Depending on the magnet system, the field of view may include the entire foot, knee, or hand, or be limited to only a couple of metacarpophalangeal (MCP) joints. The extremity magnets do not have the full breadth of pulse sequences that are available on whole-body magnets, which can limit the types of studies that may be performed.

Frequency-selective fat-saturation techniques are the most common acquisition method used for identifying edema-like marrow signal at high resolution. Frequency-selective fat saturation is also needed to increase the sensitivity of measuring contrast enhancement following intravenous injection of MRI contrast agents. Short tau inversion recovery (STIR) images provide a form of fat suppression that can make joint effusions more conspicuous, but suffer from characteristically poor SNR.

Field strength and SNR considerations.

The quality of the MR images is dependent on the magnet field strength, MRI system hardware, and imaging pulse sequence parameters. MRI systems have a number of critical components, including the main magnetic field, the gradient coils, the radiofrequency coils, and the electronics controlling them (amplifiers, pre-amplifiers). Each must be optimized for the best imaging performance. High-quality MR images must have enough spatial resolution (e.g., small enough voxel dimensions), to adequately resolve the structures of interest while maintaining an adequate SNR. The SNR of an image is directly proportional to the strength of the main magnetic field and the volume of each voxel. The SNR is also proportional to the square root of the imaging time and inversely proportional to the acquisition bandwidth, as shown in the following equation:

  • equation image

The volume of the voxel is the product of the image slice thickness and the in-plane resolution (field of view divided by the imaging matrix). The imaging matrix is the number of voxels acquired in the x- and y-axes (frequency and phase encode directions). The imaging time is determined by the number of phase encode steps in the acquisition (number of voxels acquired in the phase direction) and the number of acquisitions performed. These relationships can be used to demonstrate some differences between imaging systems.

For example, if all other factors are held constant, a magnet with a field strength of 1.5T has the potential to produce an image that has 7.5 times greater SNR than a magnet system with a field strength of 0.2T. If one relied only on imaging time to maintain SNR, the imaging time would have to be increased by 56-fold (the square of 7.5) for the low-field strength magnet to have the same SNR as the 1.5T system. Since this is not feasible, different imaging hardware and imaging parameters are used in low-field systems. Typically, at low field, lower spatial resolution (e.g., larger Vvoxel), smaller acquisition bandwidths, and different radiofrequency coils are used to increase SNR. One system (Applause) uses cryogens (liquid nitrogen) to cool the electronics to reduce electrical noise and thereby increase SNR. Despite these adjustments in hardware and pulse acquisition parameters, the low-field MRI systems are unable to obtain the SNR of high-field MRI systems for images of similar spatial resolution.

Ultimately, the image spatial resolution is dependent on the imaging parameters chosen, e.g., low-resolution protocols may be chosen for high-field magnets and vice versa, within the limitations of the hardware. When maximizing spatial resolution, a choice about the minimum acceptable SNR must be made by the operator. In making this choice, the image contrast differences between tissues also factors into the decision. If a pulse acquisition demonstrates high image contrast between the object of interest and other tissues, a lower SNR may be accepted. This may be quantified as the contrast-to-noise ratio (CNR). In general, both CNR and SNR are greater with high-field magnets compared with low-field magnets when all other parameters are held constant.

Use of intravenous contrast agents.

The field strength of the main magnetic field has an influence on the use of MRI contrast agents. MRI contrast agents work by shortening the relaxation times of tissues relative to the local concentration of the contrast agent. The most commonly used contrast agents are gadolinium chelates, which are used to shorten the spin-lattice relaxation times (T1) of tissues. At low-field strengths, intravenous contrast agents can still be utilized; however, a higher concentration of MRI contrast agent is needed to produce the same amount of image contrast enhancement as at high-field strength.

Frequency-selective fat-saturation image acquisition techniques are commonly used to improve the sensitivity of MR imaging to the effects of contrast enhancement. These techniques are especially useful for imaging tissues that contain fat, such as bone marrow and synovium, to highlight the uptake of the contrast agent. At low-field strength, these image acquisition methods cannot be used due to the very small resonant frequency differences between fat and water and the limited pulse sequence options. STIR images may be used for fat suppression on low-field MRI systems to increase the sensitivity for fluid and bone marrow edema. However, single STIR acquisitions cannot be used to improve the detection of intravenous enhancement measurements, and the images are inherently low in SNR.

Commonly performed intravenous contrast enhancement measurements include the rate of enhancement and the degree of enhancement. Due to SNR constraints, image acquisitions on low-field systems are typically much longer than those on high-field systems. Therefore, the minimum temporal resolution for measurements of tissue enhancement rates will be longer for low-field units. Since intravenously administered MRI contrast agents will diffuse from the synovium into joint fluid over a period of several minutes, rapid imaging acquisitions are desirable for the differentiation of synovium from joint fluid.

Safety considerations

  1. Top of page
  2. Introduction to the use of in-office magnetic resonance imaging (MRI)
  3. Basic MRI concepts
  4. Safety considerations
  5. Available in-office extremity MRI systems
  6. Reimbursement and obstacles to payment
  7. Predictive value of high-field MRI
  8. Review of published data regarding comparison of low- and high-field units and radiographic correlation
  9. Management considerations
  10. Summary
  11. Future research directions
  12. REFERENCES

All MR imaging systems use a very strong static magnetic field, switching gradient magnetic fields, and radiofrequency transmission to produce images. These instruments carry significant safety risks to patients and all other personnel due to the magnetic fields. It is crucial that the owner and operator of the MRI system fully understand these risks, provide proper education and training to all staff who enter the magnet area (including ancillary staff such as cleaning and emergency response personnel), and institute policies regarding access to the magnet area as well as appropriate screening of patients and personnel entering the magnet area. A full review of MRI safety is beyond the scope of this document, and readers are referred to other publications and Web sites (13–17).

The extremity magnets currently available are either permanent or superconducting magnets that are continuously operating at high magnetic fields. While superconducting magnets may be powered off in emergency situations, although at significant expense, permanent magnets always remain at the same field strength. The static magnetic fields of these magnets range from 4,000 to 20,000 times greater than the earth's magnetic field, which is 0.05 mT, or 0.5 gauss. The static magnetic field of even 0.2T permanent magnets is sufficient to attract ferromagnetic objects with substantial force to cause injury or interfere with the operation of other equipment and medical devices such as pacemakers. The static magnetic field outside the magnet bore drops quickly with distance from the magnet. It is generally believed that the safety risk of magnetic fields of ≤5 gauss (0.5 mT) is minimal. The distance from the magnet where the fringe field is <5 gauss varies with the magnet shape and design, and it is critical to know this distance so that all medical devices, ferromagnetic objects, and patients with sensitive implanted devices such as pacemakers are kept out of this area. A plan for handling patients with emergent medical conditions should be in place since some emergency equipment may not function or be safely brought into the fringe field.

Magnetic gradient fields are present at the opening and in the imaging region of the magnets during operation. Torque may be experienced by ferromagnetic and diamagnetic objects that are placed within a magnetic field or that are exposed to the switching magnetic gradient fields needed to create images. Additionally, switching magnetic gradients and radiofrequency waves have the potential to generate electrical currents within metallic objects. Some implanted medical devices or foreign objects within the body may be at risk of movement, malfunction, or damage during patient positioning or imaging; thus, proper screening for such objects is essential. Texts and Web sites are available to determine the relative risk of some of these devices and objects.

To produce the MR signal, radiofrequency energy is transmitted by the radiofrequency coil during imaging. Absorption of this energy can cause heating of tissue. Safety limits have been placed on the amount of radiofrequency delivered during imaging, to avoid excessive heating. Most manufacturers do not allow the operator to exceed these limits. However, metallic objects within or on a patient (even some tattoos) may concentrate the transmitted radiofrequency to a very small area, resulting in severe burns. Proper patient screening for such objects or risk factors is imperative.

Since every magnet system is slightly different, it is the owner's responsibility, in conjunction with the manufacturer, to develop appropriate site plans to ensure proper patient and personnel safety.

Available in-office extremity MRI systems

  1. Top of page
  2. Introduction to the use of in-office magnetic resonance imaging (MRI)
  3. Basic MRI concepts
  4. Safety considerations
  5. Available in-office extremity MRI systems
  6. Reimbursement and obstacles to payment
  7. Predictive value of high-field MRI
  8. Review of published data regarding comparison of low- and high-field units and radiographic correlation
  9. Management considerations
  10. Summary
  11. Future research directions
  12. REFERENCES

There are 4 commercially available extremity MRI systems. Their characteristics are summarized in Table 2. The Applause is a low-field MRI system that operates at 0.2T. It has a U-shaped permanent magnet. The system is cooled by liquid nitrogen to reduce electrical noise. The Artoscan-M is a low-field MRI system that operates at 0.2T. It has a slot-shaped opening for the extremity in a permanent magnet. The system does not require cryogens. The C-Scan (formerly, Artoscan-C) is a low-field MRI system that operates at 0.2T. It has a slot-shaped opening for the extremity in a permanent magnet. The system does not require cryogens. The OrthOne is a high-field MRI system that operates at 1.0T. It has a cylindrical, superconducting magnet for the extremity. The main magnet requires cryogens. To our knowledge, all of the published literature on low-field extremity magnets for MRI in RA have been performed on the Artoscan-M system or the Applause system.

Table 2. Comparison of extremity scanners*
 ApplauseArtoscan-MC-ScanOrthOne
  • *

    Not all specifications were available from materials provided by the manufacturers. FOV = field of view; NA = not available; 2-D = 2-dimensional; RF = radiofrequency; MCP = metacarpophalangeal; MTP = metatarsophalangeal.

  • Ref. 31.

  • Ref. 29.

  • §

    Ref. 42.

ManufacturerMagneVu (Carlsbad, CA)Esaote (Genoa, Italy)Esaote (Genoa, Italy)ONI (Wilmington, MA)
Field strength0.2T0.2T0.2T1.0T
Magnet typePermanentPermanentPermanentSuperconducting
CryogensLiquid nitrogenNoneNoneLiquid helium
ConfigurationU-shapeSlotSlotCylinder
Opening size15 cm × 34 cm 33.6 cm × 16 cm28-cm diameter
Purchase cost∼$145,000∼$200,000∼$250,000∼$450,000
FOV5 cm × 7.5 cm × 1 cm10–20 cm10–20 cm4–16 cm
MatrixUp to 64 × 96Up to 256 × 256Up to 256 × 256Up to 512 × 512
In-plane resolution (manufacturer limits)0.78 mm0.4 mm0.4 mm0.08 mm
Resolution (published)0.78 mm × 0.78 mm in-plane, 1 mm slice thickness0.63 mm × 0.83 mm in-plane, 1 mm slice thicknessNA0.58 mm × 0.29 in-plane, 1.5 mm slice thickness for knee examination§
Slice thickness1–10 mm2–10 mm 2-D, 0.6–10 mm 3-D2–10 mm 2-D, 0.6–10 mm 3-D2–10 mm 2-D, 0.5–10 mm 3-D
Image planesDetermined by positioningAny planeAny planeAny plane
Fringe field (0.5 mT) 0.60 m axial, 0.28 m radial0.60 m axial, 0.28 m radial1.8 m axial, 1.25 m radial
Minimum room size48 square feet100 square feet90 square feet165 square feet
Shielding requirementsNoneNoneNoneRF-shielded room to at least 80 dB
Electrical supplyNo special requirementNo special requirementNo special requirement208 volts AC 30A, 115 volts AC 20A
RF power8,000W peak 900W500W peak root mean square, 50W average
What can be imaged2–3 MCP or MTP joints or wristFingers, hand, wrist, elbow; toes, foot, ankle, kneeFingers, hand, wrist, elbow; toes, foot, ankle kneeFingers, hand, wrist, elbow; toes, foot, ankle, knee

To date, the majority of in-office extremity MRI units purchased by rheumatologists have been the 0.2T machines (Applause, C-Scan, Artoscan), although several high-field extremity systems have been acquired (OrthOne). The low-field units are small, utilize a Windows-based system, can be placed in a standard examination room, and do not require radiofrequency shielding or special power supplies, but may require liquid nitrogen cooling. The high-field extremity units are also small, but require a radiofrequency-shielded room, a dedicated power supply, and liquid helium. Patient acceptance of both types of extremity magnets is excellent due to the ease of patient positioning in contrast to conventional MRI. Images are generally limited to the wrists, MCP joints, or metatarsophalangeal (MTP) joints. Depending on the unit, the largest imaging field of view may be limited to only the second and third MCP joints (Applause), the second through the fifth MCP joints (C-Scan, Artoscan), or the wrist through the MCP joints (OrthOne). Depending on the machine and the acquisition parameters utilized, these studies may take from 15 minutes to 45 minutes. Intravenous contrast administration has not been routinely used with the low-field extremity magnets in clinical practice, in contrast to the studies on these low-field strength units reported in the peer-reviewed literature.

OrthOne (1.0T) in-office MRI scanners are being utilized by orthopedists, rheumatologists, and radiologists. Images of the larger joints, such as the elbow, knees, and ankles, can also be obtained. As with the 0.2T units, these units utilize Windows-based systems and are open MRIs. Extremity studies generally take 15–20 minutes.

According to the promotional materials, radiologic technologists are not needed to perform these studies, and no special MRI certification is necessary for the extremity systems. The images are generally transmitted to consulting radiologists who perform the interpretations on a fee-for-service basis. With MRI, however, image interpretation becomes complex since the appearance of the same condition can vary dramatically with changes in the image acquisition parameters. Thus, diagnosis using MRI may be confounded or facilitated by technical variables. These various MRI techniques must be understood and optimized for each body part, disease, and patient, and the individual determining the imaging parameters should be experienced in interpretation. Ideally, an experienced musculoskeletal radiologist and technologist would optimize these imaging parameters and monitor image quality. In the published reports, even of studies in which this expertise was utilized, interobserver variability in image interpretation was noted.

Reimbursement and obstacles to payment

  1. Top of page
  2. Introduction to the use of in-office magnetic resonance imaging (MRI)
  3. Basic MRI concepts
  4. Safety considerations
  5. Available in-office extremity MRI systems
  6. Reimbursement and obstacles to payment
  7. Predictive value of high-field MRI
  8. Review of published data regarding comparison of low- and high-field units and radiographic correlation
  9. Management considerations
  10. Summary
  11. Future research directions
  12. REFERENCES

The 0.2T extremity units cost approximately $145,000–270,000 US, and the high-field extremity units cost $450,000–500,000 US. Reimbursement for the individual studies varies depending on the insurance carrier, but generally is in the range of $350–900 US per study. The marketing materials from the manufacturers suggest that the break-even point for these machines is 1 scan per day for low-field extremity scanners and 1.5–2 scans per day for high-field extremity systems.

As with other ancillary services, health plans may choose not to reimburse for in-office MRI, and this factor should be considered prior to purchase of these units. There are presently restrictions on reimbursement in relation to the number of anatomic sites at which extremity MRI is performed at one time. For example, both wrists can be evaluated in one examination, but additional studies of the MCP or MTP joints must be performed at a second time point or they will not be reimbursed. There is concern about possible limitation by state agencies on the ability to perform in-office extremity MRI. For example, Maryland requires that only licensed radiologists perform advanced imaging procedures, such as MRI, and Rhode Island requires a certificate-of-need to perform these services. These restrictions are presently the exception, but could represent a hurdle to future reimbursement.

Predictive value of high-field MRI

  1. Top of page
  2. Introduction to the use of in-office magnetic resonance imaging (MRI)
  3. Basic MRI concepts
  4. Safety considerations
  5. Available in-office extremity MRI systems
  6. Reimbursement and obstacles to payment
  7. Predictive value of high-field MRI
  8. Review of published data regarding comparison of low- and high-field units and radiographic correlation
  9. Management considerations
  10. Summary
  11. Future research directions
  12. REFERENCES

What is the value of MRI findings of synovitis, bone marrow edema, or erosions in predicting damage that will be evident on conventional radiographs obtained at a future time? Several studies have investigated the role of MR scans in RA patients (6–9). Most of them are cross-sectional and indicate that, compared with traditional radiographs, MR scans are not only more sensitive in identifying erosions, but also allow diagnosis of them early in the course of the disease. MR scans show soft tissue abnormalities, such as synovitis, tendinitis, and bone marrow edema, that cannot be seen on conventional radiographs. However, to investigate the prognostic value of MRI findings of synovitis, bone marrow edema, and erosions in predicting radiographic damage, well-designed longitudinal studies are needed, and very few are available. Also, most of these studies have used high-field MRI (1.5T) machines and not the extremity (0.2T) MRI machines used for scanning peripheral extremity parts, such as the wrists and MCP joints.

McQueen and colleagues have studied an inception cohort of 42 patients with early RA, from presentation (median of 4 months from symptom onset) to 6 years of followup, using clinical assessments of disease activity and function as well as radiography and high-field MRI scanning of the dominant wrist (9, 12). At baseline, 45% of these patients had erosions seen on MRI compared with 15% with erosions seen on radiographs, and by year 1, 75% exhibited erosions on MRI, although erosions were seen on radiographs in only 21% of patients. McQueen et al scored the MR scans according to a locally validated scoring system and showed that the total MRI score at baseline (combining scores for erosions, bone marrow edema, synovitis, and tendinitis) was predictive of erosions on radiographs (Sharp scores) at 1, 2, and 6 years.

The MRI finding of bone marrow edema may be even more important than erosions for predicting future evidence of erosions on radiographs. Using a site-specific analysis of MR scans performed in the cohort described above, investigators in McQueen's group showed that baseline MRI-detected bone marrow edema at a specific carpal bone was highly likely to be associated with MRI erosion at that site after 1 year and 6 years (odds ratio 6.5 [95% confidence interval 2.78–18.1]) and the baseline MRI bone marrow edema score was predictive of the 6-year total Sharp score (12, 18). A model incorporating baseline MRI scores for erosion, bone marrow edema, synovitis, and tendinitis, plus the C-reactive protein (CRP) level and the erythrocyte sedimentation rate, explained 59% of the variance in the 6-year total Sharp score (R2 0.59, adjusted R2 0.44) (12). Synovitis as seen on MRI can be scored by a validated method and was a predictor of the MRI erosion score at 6 years (R2 0.15, P = 0.03), but not of the total modified Sharp score in the same cohort. This finding is similar to results of a study by Østergaard et al, who showed that MRI-detected synovitis, measured by estimation of synovial volume, was a predictor of erosions detected on MRI after 1 year (19).

Despite this observation, several caveats need to be considered. The positive predictive value of MRI scores in McQueen and colleagues' cohort was low (67%), implying that one-third of patients with a high total score on MRI at baseline did not develop radiographically evident erosions at 2 years. However, the negative predictive value was high, showing that 90% of patients with a low initial score did not develop erosions at the wrists by 2 years.

Also, the MRI findings of erosions, bone marrow edema, and synovitis may not be specific for inflammatory arthritis such as RA. In a recent study utilizing high-field MRI in assessing osteoarthritis (OA) of the hands, at least half of the patients with early OA and one-third of those with chronic OA had bone marrow edema. Erosions were even more common and were present in at least 75% of patients with early OA and 50% of patients with chronic OA. Seventy-three percent of the OA patients had excess fluid in the joint space, and gadolinium enhancement suggestive of inflammation was found in every joint studied in patients with early OA (20).

The MRI scoring system is very complex, since it includes the sum of the scores for erosions, bone marrow edema, synovitis, and tendinitis at several areas within the wrist. It is very time-consuming, needs to be applied by experts for reproducible results, and hence, is not practical to use for daily clinical studies. Simple presence or absence of bone erosion or bone marrow edema on MRI may not be predictive of long-term radiographic or functional outcome since bone marrow edema may be transient and only 26% of erosions detected on MRI progress to erosions on radiographs at 2 years. Recently, the Outcome Measures in Rheumatoid Arthritis Clinical Trials (OMERACT) group published a scoring system for high-field MRI systems (RA MRI score [RAMRIS]), which incorporates MRI features of erosion, edema, and synovitis (21). This system remains impractical for daily clinical use since it is time-consuming, complex, and exhibits significant variability in scores even with expert readers.

The reading variability with the RAMRIS scoring system can introduce a measurement error that is expressed as the smallest detectable difference (SDD), and in general, only changes greater than the SDD are considered clinically important in longitudinal studies. Ejbjerg and colleagues have compared the SDD of the RAMRIS with the Sharp/van der Heijde score (22) on radiographs in a 1-year longitudinal study (23). They assessed either 5 joints or 15 joints (wrists, MCP joints, and MTP joints) with a 0.2T low-field MRI scanner, and for determination of the Sharp/van der Heijde score they assessed 32 joints (wrists, MCP joints, proximal interphalangeal joints, and MTP joints) in 35 RA patients and 9 controls. They found that the SDD with the 5joint (MCP2–5 and dominant wrist) RAMRIS score was 2.1, compared with an SDD of 4.2 with the 15-joint RAMRIS. The SDD with the Sharp/van der Heijde score was 6.1. Defining radiographic progressors as patients with scores exceeding the SDD, more patients were designated as having progression by MRI of the dominant wrist and bilateral MCP2–5 (13 patients) than by radiography (5 patients). No difference in structural progression between MRI and radiography was noted if the dominant wrist was not included in the MRI study and only MCP and MTP joints were scanned. The authors concluded that low-field extremity MRI was more sensitive than radiographic scoring for detecting progressive joint damage.

Bird et al evaluated the progression of joint erosion over 2 years in 47 RA patients with established disease, comparing a high field-of-view MRI of the second through the fifth MCP joints with conventional bilateral hand radiographs (24). The MRIs were scored using the RAMRIS methodology and the radiographs by the Larsen score (25). In contrast to the findings reported by Ejbjerg and colleagues, more patients with progressive joint erosion were detected by bilateral hand radiography than by dominant-hand MRI. MRI did demonstrate greater sensitivity to damage progression in the MCP joints alone, but this advantage was lost when the joints of both hands were evaluated by conventional radiography. These findings suggest that, in established RA, limited-field MRI may be no better than conventional radiography in evaluating progression of joint damage.

Review of published data regarding comparison of low- and high-field units and radiographic correlation

  1. Top of page
  2. Introduction to the use of in-office magnetic resonance imaging (MRI)
  3. Basic MRI concepts
  4. Safety considerations
  5. Available in-office extremity MRI systems
  6. Reimbursement and obstacles to payment
  7. Predictive value of high-field MRI
  8. Review of published data regarding comparison of low- and high-field units and radiographic correlation
  9. Management considerations
  10. Summary
  11. Future research directions
  12. REFERENCES

There are several available studies that compare the diagnostic performance of low-field extremity MRI and conventional high-field MRI. Ejbjerg et al correlated low-field extremity MRI (0.2T) with radiography and high-field MRI (1.0T), with the latter being used as the standard (26). The dominant hand was evaluated in 37 patients who fulfilled the ACR revised diagnostic criteria for RA (27) and in 28 healthy controls. Erosions, bone marrow edema, and contrast-enhanced synovitis were compared. Two-dimensional spin-echo images with a slice resolution of 3 mm on high-field MRI were compared with 3-dimensional gradient-echo T1-weighted series on the extremity unit, using a slice resolution of 1 mm. Interobserver variability was minimal, with relatively high interclass correlation coefficients. While the comparison of synovitis disclosed that the low-field extremity unit had a reported sensitivity of 90%, specificity of 96%, and accuracy of 94%, the MRI methodology can be brought into question, since the high-field imaging was not optimized by either slice or in-plane resolution. In assessing erosions, the investigators reported a sensitivity of 94% and a specificity of 93% using high-field as a standard; however, they further stated that 318 erosions were detected with high-field MRI and 370 with low-field extremity MRI, bringing into question the false-positive rate with the low-field extremity MRI. While the investigators conceded that this was due to the higher sensitivity of the T1-weighted gradient-echo sequences, a direct comparison of optimized imaging techniques (based on in-plane and slice resolution) is not available.

Taouli and colleagues studied a cohort of 18 patients with a mean disease duration of 8 years who were receiving active therapy (28). They compared low-field extremity MRI (0.2T) with radiography and high-field 1.5T MRI examination. Only noncontrast assessment of synovial hypertrophy was performed, which deviates from prior reported standards, and they additionally assessed joint space narrowing and subjectively scored erosions. The investigators noted no statistically significant difference between the erosion and joint space narrowing scores in this small cohort. Savnik et al studied a group of 103 patients divided into 4 groups, including patients with RA for <3 years, patients with mixed connective tissue disease and psoriatic arthritis, patients with RA of >3 years' duration, and an additional cohort with arthralgia but no clinical signs of synovitis (29). A number of different pulse sequences were compared, and dynamic contrast-enhanced synovial volume was assessed. While the median volume of synovial membrane assessed on extremity MRI did not differ significantly from the volume found on high-field MRI, the high-field MRI technique was not optimized, since the investigators used an in-plane resolution of 859μ × 859μ. Also, the extremity MRI examination was performed in 2 separate scans due to the limitation of field of view (anatomic coverage). Inter- and intraobserver variability was determined only in a “randomly selected” group of patients, limiting the assessment of reproducibility.

Cimmino et al studied a group of 36 patients with RA who had a mean disease duration of 66 months (and a median age of 62.5 years) and compared these patients with a healthy control group (with a mean age of 39 years) (30). The patient group was divided into those with active RA, those with intermediate-level RA, and those whose RA was in remission. Using contrast enhancement on a 0.2T extremity unit, the investigators observed that the rate of early enhancement in the first 55 seconds, as well as the relative enhancement, was significantly higher in patients with active disease than in those with inactive disease. No reproducibility was provided, and both in-plane and slice resolution were relatively poor (938μ × 1,172μ × 5 mm). The relevance of this study to extremity MRI as presently used in clinical practice is limited since contrast enhancement is not being administered routinely in office clinical practice.

Crues and colleagues studied a cohort of 132 patients in whom MRI was administered with a 0.2T extremity unit and compared the ability of coronal T1-weighted and STIR (form of fat suppression) MRI to detect erosions, compared with radiography (31). While interobserver and intraobserver reliability were assessed as moderate (interobserver κ = 0.429), the radiographs were interpreted with knowledge of the MRI findings, which limited the comparison. The authors suggested that the inherently low SNR generated by these units is less important than CNR, which they submit is a more clinically relevant parameter. No supportive clinical data were provided. The authors further suggested that the units will permit “high resolution MR imaging,” but used a protocol with an in-plane resolution of 1,000–1,400μ, which is more than double the typical resolution used in standard examination with high-field strengths.

Lindegaard and colleagues studied a small cohort of patients with a disease duration (by clinical assessment) of <1 year, using MRI with a 0.2T extremity unit, compared with radiography and clinical assessment (32). The mean age of the 25 RA patients was 55 years. Three healthy controls, with a mean age of 46 years, were also studied. Contrast-enhanced, non–fat-suppressed T1-weighted images were scored using a subjective scale (no, slight, or marked enhancement) by a single observer (a rheumatologist), with no measure of reproducibility. The authors noted higher synovial hypertrophy grades in patients with clinical signs of joint inflammation, and 57 erosions were detected by extremity MRI, compared with 6 erosions detected on radiographs. The authors further noted that 51% of joints without clinical signs of synovitis showed “synovial hypertrophy.” Without some measure of reproducibility and a more closely matched control group, however, it is difficult to draw meaningful conclusions from the extremity MRI findings in this study.

Critical evaluation of the published results comparing extremity MRI with high-field, traditional MRI must involve a careful evaluation of MRI methodology, including the pulse sequences utilized, planes of imaging, pulse sequence parameters, and spatial (in-plane) and slice resolution. Preferably, MRIs should be interpreted by several readers, and the amount of experience reported. Despite the reported sensitivity and specificity, the study designs from a technical aspect are often variable and methodology flawed. The result may be a factitiously high accuracy of extremity MRI in disease detection, compared with high-field MRI as the standard.

Qualitative evaluation of patients with inflammatory arthritis requires careful assessment of marrow signal and cortical/subchondral bony contour, characterization of synovial enhancement patterns within the joint effusion, as well as recognition of imaging artifacts in order to avoid the false-positive MRI appearance of erosions. Artifacts that may lead to errors include the presence of inhomogeneous fat suppression, which can yield areas of abnormal signal intensity on frequency-selective fat-suppression techniques, and chemical shift misregistration, which causes factitious thinning of the subchondral bone or cartilage. In addition, “benign” cysts that are unrelated to inflammatory disease may be encountered, such as the commonly noted intraosseous ganglion cysts at the insertion of the collateral ligaments and capsule of the MCP joints. For example, Ejbjerg et al studied 28 healthy asymptomatic subjects with a mean age of 47 years, using high-field (1.0T) MRI and the OMERACT scoring system, and noted “erosion-like changes” in 2.2% of the MCP joints and 1.7% of the wrists, with “synovial-like changes” in 8.9% of MCP the joints and 9.5% of the wrists (33). These findings raise questions with regard to the ability of the inexperienced eye to distinguish benign processes at the wrist and hand joints from erosive disease, and thus render an accurate diagnosis with fewer false-positives.

The limitations in study design in the available literature comparing traditional high-field, closed-bore MRI with low-field extremity magnets, as well as the lower SNR generated from these extremity magnets, raise questions regarding the use of extremity MRI in the evaluation of patients with RA. Lower spatial and slice resolution are necessary to maintain adequate signal, which will limit the ability of such designs to be used for the early detection of disease, when more aggressive therapy may be chosen. The significance of “bone erosion” without the aid of contrast enhancement to elucidate synovitis is unclear, and less clear is the predictive value of such a lesion. Additionally, as noted in the subsequent sections of this review, the negative predictive value of a normal finding obtained with high-field machines in relation to the later development of radiographic erosions may not hold true for an extremity system that has limited capability for detecting synovitis and bone marrow edema.

The use of 3T MRI units has become increasingly popular at academic and some private imaging centers, and commercially available wrist and extremity coils are available for imaging. The SNR generated by 3T units is considerably higher than that with traditional 1.5T units, allowing for superior in-plane and slice resolution while still maintaining adequate signal. To date, however, there are no published studies that have evaluated 3T units in the assessment of RA patients, particularly in direct comparison with 1.5T units or extremity MRI.

Management considerations

  1. Top of page
  2. Introduction to the use of in-office magnetic resonance imaging (MRI)
  3. Basic MRI concepts
  4. Safety considerations
  5. Available in-office extremity MRI systems
  6. Reimbursement and obstacles to payment
  7. Predictive value of high-field MRI
  8. Review of published data regarding comparison of low- and high-field units and radiographic correlation
  9. Management considerations
  10. Summary
  11. Future research directions
  12. REFERENCES

What are the available data showing that MRI abnormalities are predictive of poorer clinical outcomes?

What data are presently available to support change in disease management based on MRI findings over and above other clinical, laboratory, and radiographic findings? Are there patients for whom the information obtained with peripheral MRI may alter the clinical decision-making process? Is there literature available to suggest that the use of high-field or peripheral MRI may alter clinical outcomes? The simple answer to these questions is no, at least according to the current published literature.

High-field MRIs have been shown to be more sensitive than both physical examination and conventional radiography at detecting erosions and synovitis in RA. However, in longitudinal studies, a significant percentage of these erosions have not progressed (9). It remains to be seen whether detection of erosions on MRI alone will have any value in guiding therapy.

To date, in studies of RA therapy using MRI data, the MRI results have not generally been used to guide treatment decisions. In one study of early RA (disease duration <12 months), comparing methotrexate with methotrexate plus intraarticular steroids, the development of erosions in the MCP joints, seen with a 1.5T MRI with contrast enhancement, over the course of a year was found to correlate with the level of synovitis assessed by MRI (10). In particular, joints without evidence of synovitis had not developed new erosions on followup MRI. Despite the value of the MRI for predicting and detecting erosions in that study, treatment decisions were driven by clinical evidence of synovitis, and not by MRI findings. It should be noted that in that study, gadolinium enhancement was used to assess the severity of synovitis, which is, as noted above, not the standard practice for low-field extremity MRI office examination.

In another small, blinded study comparing outcomes in 20 patients with early RA treated with an “induction regimen” of methotrexate with or without infliximab for 1 year, 1.5T MRIs of the second through the fifth MCP joints were evaluated for evidence of synovitis, bone marrow edema, and erosion (34). Joints were evaluated using the same protocol as in the study described above (10), including the use of intravenous gadolinium enhancement. Despite the small number of subjects in the trial, there was a significant difference between the 2 treatment groups, in both synovitis and bone marrow edema, on MRI examinations obtained as early as 14 weeks; this was sustained through 54 weeks. Radiographs showed no significant differences between the 2 groups, even after 24 months of followup. Findings on MRI did correlate with measurements of clinical outcomes, including ACR response (35), Disease Activity Score (DAS) (36), and Health Assessment Questionnaire (HAQ) score (37), although once the infliximab therapy was withdrawn at 1 year, only the HAQ scores were significantly better at 24 months in the infliximab group. While the results of this trial do suggest that MRI may provide a tool for rapid assessment of response to biologic therapy in a clinical trial and perhaps, by inference, in clinical therapy, the trial design did not allow for any assessment of how the MRI findings might be used to adjust therapy. Moreover, the differences on MRI were not reflected in differences in radiographic appearance during the 2 years of the study.

Finally, the rapid image acquisition possible with high-field MRI permits measurement of the rate of synovial uptake of gadolinium contrast agent, which has been used as a method for quantifying synovitis. In a trial of 39 patients randomized to receive leflunomide or methotrexate therapy for 4 months, MRI showed greater improvement (reduction), compared with the initial rate of enhancement, in the leflunomide group (38). However, clinical outcomes were comparable in the 2 groups over the same time period.

In conditions other than RA, MRI may be an effective element of clinical management. For example, a recent study showed that even with a mid-field 0.5T MRI of the knee, investigators were able to predict the need for arthroscopic repair of a meniscal tear, with high sensitivity and specificity (39). However, there are no published data demonstrating that use of MRI can aid in the distinction of RA from other forms of inflammatory arthritis, or aggressive RA from mild RA. Use of such data may ultimately be helpful in clinical decision-making, particularly in decisions regarding whether to initiate or continue expensive biologic therapy; existing studies, however, shed no light on these questions.

What are the available data showing that abnormalities detected using MRI are predictive of long-term disability?

Maintaining function and preventing disability are among the most important goals of arthritis treatment. If specific MRI findings could reliably predict deterioration in functional outcomes, clinicians could use these findings to guide therapy. Unfortunately, there is little evidence to date linking disability or other functional outcomes to specific findings on low-field extremity MRI.

There is a close association between the development of radiographic erosions and disability among populations of patients with RA. Because extremity MRI may be more sensitive than radiography in detecting erosions, it is possible that this imaging approach could be used to predict functional outcomes earlier and more accurately than radiography. However, there are no published studies to support this concept. In addition, the presence of radiographic erosions correlates only roughly with functional outcomes in individual patients, and the significant false-positive rate with extremity MRI (as described above) could offset its potential benefit in predicting function outcomes.

Quinn et al reported that patients with early arthritis treated with infliximab and methotrexate improved clinically and functionally compared with those taking methotrexate alone; high-field MRI evidence of synovitis mirrored these clinical and functional improvements (34). Benton and colleagues studied patients with early RA and found that baseline total MRI score and the presence of bone marrow edema on high-field MRI of the wrist predicted Short Form 36, Physical Function (PF–SF-36) scores (40) at 6 years (18). In fact, 16% of the PF–SF-36 score was explained by the baseline total MRI score, and 22% of the PF–SF-36 score was explained by the presence of bone marrow edema. However, HAQ scores at 6 years were not predicted by MRI results, and baseline Ritchie Articular Index (41) and baseline HAQ score predicted 6-year HAQ score as well as did MRI (20% of the 6-year HAQ score was explained by these other baseline assessments). The authors noted that the best predictor of 6-year function was a regression model that included 1) bone marrow edema by MRI, 2) CRP level, 3) DAS, 4) HAQ score, and 5) modified Sharp score.This model predicted 23% of the 6-year PF–SF-36 score. Thus, although this study showed correlations between certain functional outcomes and baseline high-field extremity MRI findings, the ability to predict outcomes was modest. In addition, it is unlikely that a clinician using extremity MRI in the office will utilize the radiographic and MRI scoring systems or the regression model described in that report. Importantly, the study was performed prior to the introduction of anti–tumor necrosis factor therapies.

Summary

  1. Top of page
  2. Introduction to the use of in-office magnetic resonance imaging (MRI)
  3. Basic MRI concepts
  4. Safety considerations
  5. Available in-office extremity MRI systems
  6. Reimbursement and obstacles to payment
  7. Predictive value of high-field MRI
  8. Review of published data regarding comparison of low- and high-field units and radiographic correlation
  9. Management considerations
  10. Summary
  11. Future research directions
  12. REFERENCES

The benefits of low-field strength extremity MRI for the diagnosis and management of rheumatoid arthritis are still being elucidated. Our review raises the following issues:

  • 1
    The literature assessing the utility of peripheral joint MRI has used high-field, not low-field extremity MRI; therefore, actual sensitivity, specificity, and predictive value of the low-field scanners available for the practicing rheumatologists are not known.
  • 2
    The high sensitivity of MRI for the detection of erosions may be at least partially offset by a lack of specificity; a significant false-positive rate is suggested by the finding that many “erosions” detected by MRI do not progress to radiographic erosions over many years. The one published study evaluating the predictive value of a single MRI erosion showed that only 26% of these became radiographically evident erosions at 2 years (9).
  • 3
    The published literature on the importance of erosions in RA and their correlation with functional decline is based on radiographic studies, not MRI series, and the predictive value of erosions detected by MRI may be significantly less than is assumed (especially for those erosions that are never identified by radiography even after several years).
  • 4
    Ideally, scanning would be performed only when the results would provide information that is otherwise unavailable and would affect management; for example, the patient and clinician are unlikely to be helped by peripheral MRI if the patient is receiving maximal therapy, with no evidence of active disease, or if clinical disease activity is significant and treatment change would be indicated regardless of the MRI findings.
  • 5
    The marginal benefit of low-field extremity MRI above and beyond standard measures of disease activity and severity (including history, physical examination, selective laboratory testing, and radiography of the hands and wrists) has not been rigorously evaluated in studies published to date. In fact, at least one study suggests that high-field MRI is no better than bilateral hand radiography in detecting progression of joint damage over 2 years, presumably because more joints are surveyed by radiography.
  • 6
    Imaging may be helpful in select patients when the history, physical examination, routine laboratory testing, and standard radiographic imaging are inadequate to enable a rational decision regarding management; however, the indications for such imaging remain uncertain. A negative result in a patient with mild RA disease activity might provide confidence for use of less aggressive treatment, but to date there have been no studies to validate this hypothesis using low-field extremity MRI.
  • 7
    There is currently no consensus regarding when high-field MRI should be ordered for the diagnosis and management of RA. Due to the limitations of study design and the limited number of studies, it is even more difficult to establish clinical indications for low-field extremity scanning.

Future research directions

  1. Top of page
  2. Introduction to the use of in-office magnetic resonance imaging (MRI)
  3. Basic MRI concepts
  4. Safety considerations
  5. Available in-office extremity MRI systems
  6. Reimbursement and obstacles to payment
  7. Predictive value of high-field MRI
  8. Review of published data regarding comparison of low- and high-field units and radiographic correlation
  9. Management considerations
  10. Summary
  11. Future research directions
  12. REFERENCES

Many important issues relevant to the optimal utilization of MRI remain unanswered and should be addressed in prospective research studies. Some of the questions include the following:

  • 1
    In patients with RA, what is the long-term significance, for example, in terms of functional disability, of periarticular erosions demonstrated by MRI that were not apparent on radiography? Is there any significance to such erosions seen on high-field (e.g., 1.5T) systems or with special techniques that would not have been demonstrated by low-field machines (e.g., 0.2T)?
  • 2
    Is it possible to differentiate RA from spondylarthropathies (e.g., enthesitis) based on extremity MRI findings? If so, can comparable differentiation be made using low-field as compared with high-field systems, and would this finding aid in patient management?
  • 3
    In patients with clinically apparent inflammatory synovitis, is there any utility of sequential MRI assessment with regard to prognosis or response to therapy? If so, how comparable is imaging at different field strengths?
  • 4
    For the above questions, are additional refinements in MRI (e.g., gadolinium enhancement, higher-resolution imaging, angiographic techniques) of value in providing an answer? If so, what machines are capable of performing these techniques?
  • 5
    What is the test–retest reproducibility for the detection and grading of MRI erosions and edema-like marrow abnormalities using low-field and high-field MRI?
  • 6
    What is the marginal benefit of low-field extremity MRI for individual patients compared with standard clinical and radiographic assessments? Highly active, aggressive inflammatory arthritis warrants aggressive treatment regardless of MRI findings; therefore, the most helpful design might be a study that analyzes functional outcomes in patients with mild-to-moderate early arthritis, for whom treatment is based on the results of low-field extremity MRI. Such a study could compare a realistic model of “usual care” with aggressive, MRI-driven therapy over a number of years. Functional outcomes could be correlated with the results of specific low-field extremity MRI findings, including bone marrow edema, synovitis, erosions, and a combination of these findings, to determine which is most useful. A study of functional outcomes that included both optimized high-field and low-field extremity MRI with multiple observers would be invaluable in establishing and comparing the utility of these imaging techniques.

MRI techniques are changing at a rapid pace. It is likely that in the future—perhaps in the near future—new and better information will become available that refines what we know about the diagnostic and prognostic implications of MRI abnormalities, how these abnormalities correlate with histopathologic findings, and how well they predict long-term clinical outcomes. Conclusions regarding the clinical utility of low-field MRI in patients with rheumatoid arthritis and other rheumatic diseases must be revisited and revised as this additional information becomes available.

REFERENCES

  1. Top of page
  2. Introduction to the use of in-office magnetic resonance imaging (MRI)
  3. Basic MRI concepts
  4. Safety considerations
  5. Available in-office extremity MRI systems
  6. Reimbursement and obstacles to payment
  7. Predictive value of high-field MRI
  8. Review of published data regarding comparison of low- and high-field units and radiographic correlation
  9. Management considerations
  10. Summary
  11. Future research directions
  12. REFERENCES
  • 1
    Kirwan JR. Conceptual issues in scoring radiographic progression in rheumatoid arthritis [review]. J Rheumatol 1999; 26: 7205.
  • 2
    Sharp JT, Young DY, Bluhm GB, Brook A, Brower AC, Corbett M, et al. How many joints in the hands and wrists should be included in a score of radiologic abnormalities used to assess rheumatoid arthritis? Arthritis Rheum 1985; 28: 132635.
  • 3
    Lipsky PE, van der heide DM, St. Clair EW, Furst DE, Breedveld FC, Kalden JR, et al, and the Anti-Tumor Necrosis Factor Trial in Rheumatoid Arthritis with Concomitant Therapy Study Group. Infliximab and methotrexate in the treatment of rheumatoid arthritis. N Engl J Med 2000; 343: 1594602.
  • 4
    Bathon JM, Martin RW, Fleischmann RM, Tesser JR, Schiff MH, Keystone EC, et al. A comparison of etanercept and methotrexate in patients with early rheumatoid arthritis [published erratum appears in N Engl J Med 2001;344:240]. N Engl J Med 2000; 343: 158693.
  • 5
    Keystone EC, Kavanaugh AF, Sharp JT, Tannenbaum H, Hua Y, Teoh LS, et al. Radiographic, clinical, and functional outcomes of treatment with adalimumab (a human anti–tumor necrosis factor monoclonal antibody) in patients with active rheumatoid arthritis receiving concomitant methotrexate therapy: a randomized, placebo-controlled, 52-week trial. Arthritis Rheum 2004; 50: 140011.
  • 6
    Forslind K, Larsson EM, Johansson A, Svensson B. Detection of joint pathology by magnetic resonance imaging in patients with early rheumatoid arthritis. Br J Rheumatol 1997; 36: 6838.
  • 7
    Foley-Nolan D, Stack JP, Ryan M, Redmond U, Barry C, Ennis J, et al. Magnetic resonance imaging in the assessment of rheumatoid arthritis: a comparison with plain film radiographs. Br J Rheumatol 1991; 30: 1016.
  • 8
    Gilkeson G, Polisson R, Sinclair H, Vogler J, Rice J, Caldwell D, et al. Early detection of carpal erosions in patients with rheumatoid arthritis: a pilot study of magnetic resonance imaging. J Rheumatol 1988; 15: 13616.
  • 9
    McQueen FM, Benton N, Crabbe J, Robinson E, Yeoman S, McLean L, et al. What is the fate of erosions in early rheumatoid arthritis? Tracking individual lesions using x-rays and magnetic resonance imaging over the first two years of disease. Ann Rheum Dis 2001; 60: 85968.
  • 10
    Ostergaard M, Hansen M, Stoltenberg M, Jensen KE, Szkudlarek M, Pedersen-Zbinden B, et al. New radiographic bone erosions in the wrists of patients with rheumatoid arthritis are detectable with magnetic resonance imaging a median of two years earlier. Arthritis Rheum 2003; 48: 212831.
  • 11
    Conaghan PG, O'Connor P, McGonagle D, Astin P, Wakefield RJ, Gibbon WW, et al. Elucidation of the relationship between synovitis and bone damage: a randomized magnetic resonance imaging study of individual joints in patients with early rheumatoid arthritis. Arthritis Rheum 2003; 48: 6471.
  • 12
    McQueen FM, Benton N, Perry D, Crabbe J, Robinson E, Yeoman S, et al. Bone edema scored on magnetic resonance imaging scans of the dominant carpus at presentation predicts radiographic joint damage of the hands and feet six years later in patients with rheumatoid arthritis. Arthritis Rheum 2003; 48: 181427.
  • 13
    Shellock FG, Kanal E. Magnetic resonance: bioeffects, safety and patient management. 2nd ed. New York: Lippincott-Raven; 1996.
  • 14
    Shellock FG. Reference manual for magnetic resonance safety, implants and devices. Los Angeles: Biomedical Research Publishing; 2005.
  • 15
    Shellock FG. Magnetic resonance procedures: health effects and safety. Boca Raton: CRC Press; 2001.
  • 16
  • 17
  • 18
    Benton N, Stewart N, Crabbe J, Robinson E, Yeoman S, McQueen FM. MRI of the wrist in early rheumatoid arthritis can be used to predict functional outcome at 6 years. Ann Rheum Dis 2004; 63: 55561.
  • 19
    Ostergaard M, Hansen M, Stoltenberg M, Gideon P, Klarlund M, Jensen KE, et al. Magnetic resonance imaging–determined synovial membrane volume as a marker of disease activity and a predictor of progressive joint destruction in the wrists of patients with rheumatoid arthritis. Arthritis Rheum 1999; 42: 91829.
  • 20
    Tan AL, Grainger AJ, Tanner SF, Shelley DM, Pease C, Emery P, et al. High-resolution magnetic resonance imaging for the assessment of hand osteoarthritis. Arthitis Rheum 2005; 52: 235565.
  • 21
    Ostergaard M, Edmonds J, McQueen F, Peterfy C, Lassere M, Ejbjerg B, et al. An introduction to the EULAR-OMERACT rheumatoid arthritis MRI reference image atlas. Ann Rheum Dis 2005; 64 Suppl 1: i37.
  • 22
    Van der Heijde DM, van Riel PL, Nuver-Zwart IH, Gribnau FW, van de Putte LB. Effects of hydroxychloroquine and sulphasalazine on progression of joint damage in rheumatoid arthritis. Lancet 1989; 1: 10368.
  • 23
    Ejbjerg BJ, Vestergaard A, Jacobsen S, Thomsen HS, Ostergaard M. The smallest detectable difference and sensitivity to change of magnetic resonance imaging and radiographic scoring of structural joint damage in rheumatoid arthritis finger, wrist, and toe joints: a comparison of OMERACT rheumatoid arthritis magnetic resonance imaging score applied to different joint combinations and the Sharp/van der Heijde radiographic score. Arthritis Rheum 2005; 52: 23006.
  • 24
    Bird P, Kirkham B, Portek I, Shnier R, Joshua F, Edmonds J, et al. Documenting damage progression in a two-year longitudinal study of rheumatoid arthritis patients with established disease (the DAMAGE study cohort): is there an advantage in the use of magnetic resonance imaging compared with plain radiography? Arthritis Rheum 2004; 50: 13839.
  • 25
    Larsen A, Dale K, Eek M. Radiographic evaluation of rheumatoid arthritis and related conditions by standard reference films. Acta Radiol 1977; 18: 48191.
  • 26
    Ejbjerg B, Narvestad E, Jacobsen S, Thomsen HS, Ostergaard M. Optimized low cost, low field dedicated extremity MRI is highly specific and sensitive for synovitis and bone erosions in rheumatoid arthritis wrist and finger joints: a comparison with conventional high-field MRI and radiography. Ann Rheum Dis 2005; 64: 12807.
  • 27
    Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988; 31: 31524.
  • 28
    Taouli B, Zaim S, Peterfy C, Lynch JA, Stork A, Guermazi A, et al. Rheumatoid arthritis of the hand and wrist: comparison of three imaging techniques. AJR Am J Roentgenol 2004; 182: 93743.
  • 29
    Savnik A, Malmskov H, Thomsen HS, Bretlau T, Graff LB, Nielsen H, et al. MRI of the arthritic small joints: comparison of extremity MRI (0.2 T) vs high-field MRI (1.5 T). Eur Radiol 2001; 11: 10308.
  • 30
    Cimmino MA, Innocenti S, Livrone F, Magnaguagno F, Silvestri E, Garlaschi G. Dynamic gadolinium-enhanced magnetic resonance imaging of the wrist in patients with rheumatoid arthritis can discriminate active from inactive disease. Arthritis Rheum 2003; 48: 120713.
  • 31
    Crues JV, Shellock FG, Dardashti S, James TW, Troum OM. Identification of wrist and metacarpophalangeal joint erosions using a portable magnetic resonance imaging system compared to conventional radiographs. J Rheumatol 2004; 31: 67685.
  • 32
    Lindegaard H, Vallo J, Horslev-Peterson K, Junker P, Ostergaard M. Low field dedicated magnetic resonance imaging in untreated rheumatoid arthritis of recent onset. Ann Rheum Dis 2001; 60: 7706.
  • 33
    Ejbjerg B, Narvestad E, Rostrup E, Szkudlarek M, Jacobsen S, Thomsen HS, et al. Magnetic resonance imaging of wrist and finger joints in healthy subjects occasionally shows changes resembling erosions and synovitis as seen in rheumatoid arthritis. Arthritis Rheum 2004; 50: 1097106.
  • 34
    Quinn MA, Conaghan PG, O'Connor PJ, Karim Z, Greenstein A, Brown A, et al. Very early treatment with infliximab in addition to methotrexate in early, poor-prognosis rheumatoid arthritis reduces magnetic resonance imaging evidence of synovitis and damage, with sustained benefit after infliximab withdrawal: results from a twelve-month randomized, double-blind, placebo-controlled trial. Arthritis Rheum 2005; 52: 2735.
  • 35
    Felson DT, Anderson JJ, Boers M, Bombardier C, Furst D, Goldsmith C, et al. American College of Rheumatology preliminary definition of improvement in rheumatoid arthritis. Arthritis Rheum 1995; 38: 72735.
  • 36
    Prevoo ML, van ‘t Hof MA, Kuper HH, van Leeuwen MA, van de Putte LB, van Riel PL. Modified disease activity scores that include twenty-eight–joint counts: development and validation in a prospective longitudinal study of patients with rheumatoid arthritis. Arthritis Rheum 1995; 38: 448.
  • 37
    Fries JF, Spitz PW, Kraines RG, Holman HR. Measurement of patient outcome in arthritis. Arthritis Rheum 1980; 23: 13745.
  • 38
    Reece RJ, Kraan MC, Radjenovic A, Veale DJ, O’Connor PJ, Ridgway JP, et al. Comparative assessment of leflunomide and methotrexate for the treatment of rheumatoid arthritis, by dynamic enhanced magnetic resonance imaging. Arthritis Rheum 2002; 46: 36672.
  • 39
    Vincken PW, ter Brakk BP, van Erkell AR, de Rooy TP, Mallens WM, Post W, et al. Effectiveness of MRI imaging in selection of patients for arthroscopy of the knee. Radiology 2002; 223: 73946.
  • 40
    Ware JE Jr, Sherbourne CD. The MOS 36-item Short-Form health survey (SF-36). I. Conceptual framework and item selection. Med Care 1992; 30: 47383.
  • 41
    Ritchie DM, Boyle JA, McInnes JM, Jasani MK, Dalakos TG, Grieveson P, et al. Clinical studies with an articular index for the assessment of joint tenderness in patients with rheumatoid arthritis. QJM 1968; 37: 393406.
  • 42
    Beattie KA, Boulos P, Pui M, O'Neill J, Webber CE, Adachi JD. Abnormalities identified in the knees of asymptomatic volunteers using peripheral magnetic resonance imaging. Osteoarthritis Cartilage 2005; 13: 1816.