NCBI Bookshelf. A service of the National Library of Medicine, National Institutes of Health.
Chung M, Dahabreh IJ, Hadar N, et al. Emerging MRI Technologies for Imaging Musculoskeletal Disorders Under Loading Stress [Internet]. Rockville (MD): Agency for Healthcare Research and Quality (US); 2011 Nov. (Comparative Effectiveness Technical Briefs, No. 7.)
Emerging MRI Technologies for Imaging Musculoskeletal Disorders Under Loading Stress [Internet].
Show detailsDescription of Stress-Loading MRI
This section addresses the operating principles and potential benefits and harms associated with stress-loading magnetic resonance imaging (MRI). In addition, we discuss the availability and cost of the relevant imaging technologies and the requirements for their use. Findings are based on multiple data sources, including one-to-one discussions with Key Informants (KIs), our review of the literature (including grey literature sources), a review of device manufacturers’ and health insurance companies’ Web sites, and searches of the Food and Drug Administration’s (FDA) Manufacturer and User Facility Device Experience (MAUDE) database and Section 510(k) of the Food, Drug, and Cosmetic Act database.
What are the operating principles of stress-loading MRI, and what are the potential benefits and harms associated with its use?
We identified several configuration or imaging acquisition features that we considered relevant to produce an accurate description of the capabilities and limitations of the diverse MRI imaging systems designed to improve accuracy by imaging joints under physiologic conditions. We focused on three distinct features of these emerging MRI technologies and applications (see Figure 3): (1) the ability to choose alternative positioning, (2) the ability to perform imaging under stress-loading conditions; and (3) having an open or semi-open configuration. Specific devices may incorporate one or more of these features. Figure 3 details the possible combinations of these features among different devices in a Venn diagram. Given the multitude of device configurations in the studies we reviewed, it is likely that the diagram is not comprehensive. However, it is sufficient to categorize the MRI machines that were used in the reviewed studies, and was therefore helpful in organizing studies and comparisons across device categories. Theoretically, all three features could appear in both high- and low-field strength MRI scanners; however, open MRI scanners have typically been of low magnetic field strengths.
More specifically, the operational definitions for these three features are as follows:
- Positioning. Special devices or placement techniques that allow imaging of the patient in postures other than the typical supine position, such as placing the spine or joint in the position of pain or anatomic abnormality through flexion or extension. Positional image acquisition can be either static or dynamic (during movement between different positions), or conducted under weight-bearing or non-weight-bearing conditions.
- Stress-loading. Devices that enable imaging in weight -bearing positions or simulate gravity (for example through axial loading, a technique that compresses the body along the joint of interest) in open, semi-open, or conventional scanners.
- Open or semi-open configuration. Any MRI scanner that does not require the patient’s body to be placed in a closed (typically cylindrical) bore. With appropriate mechanical modifications, such devices can be used to obtain images under stress-loading or weight-bearing conditions, or positions of pain or anatomic abnormality through flexion or extension.
Seven categories, as defined by the seven distinct regions of the three overlapping feature circles, are labeled A through G. In the present Technical Brief, we include all studies falling under areas A to D (stress loading). Given the diversity of available technologies and the large number of potential modifications to existing systems, it is infeasible to review all studies of open MRI devices or all studies where non-supine patient positioning was attempted in a closed scanner (via restraints or voluntary flexion/extension, etc.). We excluded studies that used open MRIs without using positioning or stress loading (area E), and studies that used devices or protocols falling under areas F and G because they do not provide information on stress-loading MRI and as such were considered outside the scope of this Technical Brief.a
All MRI is based on the same principles of physics to generate images. For stress-loading MRI devices, based on a review of the literature and KI input, we compiled a list of features relevant for comparison in evaluating patients with musculoskeletal symptoms (Table 1). We discuss these features of emerging MRI technologies as follows.
Table 1
Features for stress-loading MRI.
Imaging Under Stress-Loading and in Symptomatic Positions (Static or Dynamic)
The ability to reproduce physiologic stresses via weight bearing or simulated weight bearing may increase the ability to diagnose pathologic conditions that anatomically manifest only under stress (physiologic or supraphysiologic). In theory, imaging acquisition in stress-loading conditions can be performed in either high- or low-field strength MRI systems with open, semi-open, or closed configurations. In practice, techniques to obtain stress-loading images are more limited in a closed-bore MRI scanner due to the lack of space. Thus, stress-loading imaging in conventional scanners is usually obtained via the use of axial-loading devices to “simulate” gravity in the supine position. In contrast, there is a diversity of technologies available, as well as a large number of potential modifications that can be made, to existing open or semi-open scanners to obtain stress-loading images.
Open scanners provide extra (patient and working) space that allows for the acquisition of images under physiologic stress. Most of these systems are open laterally, and patients are typically placed in a supine (non-weight-bearing) position. We identified one open scanner device, with a tilting design, that can perform scans both in the supine position and in the upright weight-bearing position, and two vertically open (upright) scanners that allow positional (flexion, extension), weight-bearing (upright and sitting) imaging. Vertically open (upright) scanners allow the acquisition of scans in several different positions, including flexion and extension views of the cervical and lumbar spine, the knees, or the shoulders. The ability to obtain MRI imaging at joint positions physiologically similar to those examined during the clinical examination, such as placing the spine or joint in the position of pain or anatomic abnormality through flexion or extension, may reveal additional abnormalities. For those scanners which allow for dynamic imaging, special software is required.
The primary disadvantage of open, weight-bearing MRIs is their low field strength (existing systems have a maximum field strength of 0.6 Tesla [T]), which is associated with lower image quality (also see the Field Strength and Image Quality section).
Another approach used to obtain imaging under stress is through the use of specialized devices that exert force on specific joints while the patient is positioned in a conventional (closed-bore) MRI scanner. Most commonly, commercially available or custom-made devices are used to obtain MRI images of the spine under simulated gravity, but imaging of other joints, such as the knee, is also possible. For spinal imaging, gravity is typically simulated using a compressive system comprised of a vest worn by the patient over the shoulders and upper chest attached to a footplate (via cables or other means) against which the patient’s feet are braced. The axial force, in theory, reproduces the effects of upright posture. The force applied is usually 50 percent or less of that produced by the patient’s total weight. The major disadvantage of axial-loading MRI is that it requires non loaded imaging to be obtained prior to the loaded scan to rule out certain pathologic conditions that would render the scan unsafe, and the axial-loading equipment may induce or worsen pain or neurological symptoms. Some of the KIs expressed concern that some patients may not be able to complete weight -bearing, stress loading, or positional imaging due to the development of neurological symptoms, induced by the stress on the imaged joints during image acquisition.
Positional imaging is not possible in the commercially available axial-loading devices, as they require the patient to maintain an extended position in order for the compressive force to be transferred to the spine. However, we did identify descriptions of weight-bearing imaging where compression was applied while the knees were flexed (in most cases, to obtain images of the knee joints under physiologic conditions).
A KI suggested that imaging of the joints in desired positions may be obtained virtually with any imaging device, provided appropriate mechanical modifications or “ingenuity” in patient placement. In the Evidence Map subsection of this section of the report, we describe in more detail the specific modifications of existing devices encountered in the published literature. Indeed, our literature search identified many studies where commercially available MRI devices were mechanically modified to obtain imaging under stress-loading conditions that were considered desirable.
Field Strength and Image Quality
While no professional society defines “low-field strength,” the scientific literature indicates that the commonly used cutoff point for low-field strength is below 0.5 T, 0.5 to 1.0 T for mid-field strength, and greater than 1.0 T for high-field strength.38 In the studies we reviewed, open MRI scanners typically had field strengths less than 1.0 T and that open, weight-bearing MRI devices had a maximum field strength of 0.6 T.
There was agreement among all KIs and external information sources (medical physics literature, Internet Web sites) that for MRI to be effectively used for clinical decision making, the most important feature is the ability to generate images of sufficiently high quality. All sources agreed that magnetic field strength is generally the major determinant of image quality, although it should be noted that improved image quality may not directly translate to improved clinical or diagnostic utility of the scans.39 Both clinicians as well as payer representatives expressed concerns about the image quality of all devices with low magnetic field strength and questioned whether these scanners produce images of adequate quality to be used for diagnostic purposes. Among the factors that determine image quality, signal-to-noise ratio, contrast-to-noise ratio, and artifacts are three characteristics that are clearly influenced by field strength. However, which of these parameters has the most relevance to clinical diagnosis remains unanswered. 39 Information from manufacturers’ materials stated that since all available scanners have achieved the minimum regulatory technical requirements for image quality, therefore each device’s technical specifications should be considered adequate for clinical use. An industry KI stated that the use of specialized receiver coils can compensate for the low field strength limitation in terms of image quality. We found that a variety of receiver coils were used based on the configuration of the MRI device but did not find supporting information on this claim in the literature.
Image Acquisition Speed
Gradient coils that can generate high gradient strengths and slew rates are required to produce high imaging speeds and improved image quality.40 The ability to obtain images faster may increase image quality (by reducing motion artifacts) and patient comfort (by reducing acquisition times). Patient comfort may influence patient preferences for the choice of a specific diagnostic modality. Furthermore, position or loading stress-related discomfort may influence the quality of the diagnostic information obtained from an MRI scan as patients may not be able to maintain a stable position during image acquisition (because of pain or other neurological symptoms) and thus create motion artifacts.
Patient Safety
Of particular concern for emerging technologies is whether their use is associated with risks to patients. Based on KI interviews and our review of the literature, there appear to be no additional serious safety issues associated with stress-loading MRI per se as compared with conventional MRI technologies. However, the standard limitations that apply to MRI in conventional scanners apply to open and stress-loading MRIs as well: patients with metal implants, including those with surgical implants, intracranial aneurysm clips or pacemakers may not be able to be scanned; there is a potential for adverse events from the use of paramagnetic contrast agents; and hearing protection while undergoing MRI is required. Information related to MRI safety issues and guidance on best practices for ensuring patient and provider safety are provided by the American College of Radiology,41 the International Society for Magnetic Resonance in Medicine (available at: http://www.ismrm.org/), the FDA,42 and the ECRI Institute.43
We searched the FDA MAUDE database for reported adverse events associated with the use of any of the specific stress-loading MRI devices we identified through KI input and our MEDLINE and grey literature searches. We identified two MAUDE Product Problem Report documents (dated July 2006, October 2007; May 2009) regarding patients’ safety associated with the Fonar UPRIGHT Multi-Position MRI device. These two reports discussed the occurrence of skin burns due to electrical contacts on the patient’s bed, but indicated that these were cases of operator error and not a defect in the device. The device contains caution labels on the bed and instructions in the user manual indicating that contact with the skin may induce radio frequency burns during scanning and instructs that the contacts be covered by the provided safety covers or bed cushion.44,45 Another report discussed a solder joint failure in the motion control computer power supply of the device. That report led to action on the part of the company in order to replace the power supply and its harness with components that had been modified to correct the problem for all devices on the market.46
As previously mentioned, axial-loading devices apply a compressive force to the spine and are indicated for those patients in whom a diagnosis may not have yet been established (for example patients with “nonspecific spine pain” referred for MRI imaging). It is important that alternative imaging methods are used in these patients prior to capture of the axially loaded image to rule out fracture or neoplastic disease, as the additional force may cause spinal cord compression.
Bore Configuration
The majority of KIs mentioned that a substantial proportion of patients may avoid undergoing an MRI scan due to claustrophobia, and that some obese patients may not be able to undergo conventional MRI. These are both well recognized problems in the literature and appears to affect both children and adults. There is evidence that 4 to 30 percent of patients experience anxiety-related reactions while undergoing MRI.23 Furthermore, due to the global rise in obesity, the need for wider-bore MRI systems may increase.23,47–49 Some of the KIs suggested that scanning in open configuration scanners (including vertically open devices) may alleviate patient fears as well as accommodate larger patients. 23 Therefore, open MRI configurations, which have a wide gap between their magnets, may allow more patients to undergo needed imaging exams. Other open MRI configurations include semi-open, extremity-specific MRI systems (that is, systems that image specific joints such as the knee, elbow, or wrist). Such systems are rarely used in stress-loading applications, although they could in theory, and we found no published studies of such uses. There are, however, a significant number of studies of non-stress loading applications of extremity specific MRIs, which we have listed in Appendix F.
KIs also mentioned that a new, more expansive MRI system, called wide-bore MRI, was recently introduced, and serves the needs of obese and claustrophobic patients requiring larger imaging systems.50–52 Wide-bore MRI (sometimes referred to as “open-bore,” not to be confused with “open MRI”) is similar in configuration and strength (1.5 or 3T) to that of a conventional MRI but with a larger space in the cylindrical bore. However, we did not find research employing wide-bore MRI devises in stress-loading applications.
We summarize the aforementioned theoretical advantages and disadvantages of competing MRI technologies in Table 2.
Table 2
Theoretical advantages and disadvantages of conventional and emerging MRI technologies.
Are stress-loading MRIs used for routine clinical assessments or for research purposes?
In searching studies for evidence map, we did not find any published data that addressed this question. However, the general consensus of the clinical KIs was that stress-loaded MRIs are not commonly used for the diagnosis or management of musculoskeletal disorders, and should still be considered experimental. In agreement with this, most insurers’ policies that we reviewed consider stress-loading MRI as investigational and not medically necessary for any indication. Many insurers’ policies were based on literature reviews and referenced the systematic Health Technology assessment of positional MRI conducted on behalf of the State of Washington.24 Frequently, insurers reported conducting regularly updated literature searches to support their policy decisions, and, in the vast majority of cases, the evidence was considered as insufficient to demonstrate clinical utility beyond that of conventional MRI. We identified only one insurer that considered “open MRI units of any configuration, including MRI units that allow imaging when standing (stand-up MRI) or when sitting, to be an acceptable alternative to standard closed MRI units.” However, this insurer also considered “repeat MRI scans in different positions (such as flexion, extension, rotation and lateral bending) and when done with and without weight-bearing to be experimental and investigational.”53 These policies probably reflect that stress loading MRI technologies are relatively at an early stage of the diagnostic test development process (see the Evidence Map subsection of this section and the Next Steps section of this report).
To which populations and for what indications might stress-loading MRI apply? Is stress-loading MRI being proposed as a replacement, triage, or add-on test?
There was substantial divergence of opinion between the clinical experts (orthopedic surgeon and radiologist) and payers, our industry sources regarding how stress-loading MRI should be used in comparison to conventional MRI, when, and for whom. In general, clinical KIs indicated that there was not enough evidence to answer these questions, and that, in their (clinician and payer KIs’) experience, whether these devices were used for triaging patients to other imaging modalities, as replacements of or add-ons to conventional MRI, varied greatly in clinical practice. In contrast, industry input suggested that the devices could be used as replacement tests for conventional MRI for imaging the cervical and lumbar spine.
As noted earlier, conventional imaging must be conducted prior to applying simulated loads to rule out pathologic conditions that would render the scan unsafe. As such, applications of axially-loaded tests are de facto “add-on tests,” and studies of axial-loading devices tend to be comparative in design (i.e., comparing preloaded MRI with loaded MRI images). The FDA 510(k) document of the axially loaded MRI devices states, “Ideally, the examination is performed directly after the basic unloaded investigation and thus decided by the radiologist.”
Who are the current (major) manufacturers of stress-loading MRI devices? What is the current FDA clearance status of these MRI devices?
We identified the following three manufacturers of weight-bearing MRI devices (listed chronologically by FDA clearance date): (1) Signa SP/2 (General Electric Medical Systems) [510(k) # K893509], (2) Indomitable MRI Scanner (Fonar Corporation) [510(k) # K002490] (later brand name Upright MRI); and (3) Esaote S.p.A G-SCAN [510(k) # K042236] All three devices were cleared by the FDA on a “substantially equivalent” basis with predicate MRI scanners.
We identified the following two commercially available axial-loading devices, commonly referred to as medical compression devices (listed chronologically by FDA clearance date): (1) DynaWell L-spine compression device [510(k) # K992120]; and (2) Choy Compression Frame (Choy Medical Technologies) [510(k) # K070968].The former device was cleared by the FDA on a “substantially equivalent” basis for the indication for use stated as an accessory for axial compression of the lumbar spine in computerized tomography (CT) and MRI. The later device was considered as “substantially equivalent” to the DynaWell device.
Approximately how many and of what kind are facilities currently providing stress-loading MRI testing in the United States?
According to the KI affiliated with Fonar, there are about 70 American College of Radiology accredited MRI providers equipped with the Upright MRI, and approximately 140 Upright MRI scanners are currently installed worldwide (mostly in the United States). This was according to a Fonar press release from late 2009.54 We could not obtain estimates for the number of Sign a SP/2 devices available.
The Signa device was marketed as an interventional and intra operative magnetic MRI (with a 56-cm-wide vertical gap, allowing access to the patient and permitting the execution of MRI-guided interventional procedures). Many investigators had modified this device and used it to provide weight-bearing MRI testing. Our KI affiliated with GE Healthcare indicated that, to the best of his knowledge, the Signa SP/2 was still on the market however external information indicated that the device is no longer marketed by GE Healthcare in the USA.
What kinds of training, certification, and staffing are required to operate stress-loading MRI or to interpret its images?
There is no specific mandatory accreditation for the operation of any of the stress-loading MRI devices, including, open, upright, and extremity MRIs. Specialized personnel would have to become proficient with the operational software and in the patient positioning platforms. One KI affiliated with a weight-bearing MRI manufacturer indicated that the manufacturer provided training for positioning patients and scanning protocols with the installation of a new scanner.
In most cases, a board-certified radiologist is required to interpret MRI images. KIs indicated that images generated by low field strength systems may require more “experienced” readers compared with those from high field strength systems (e.g., 1.5T systems, which are currently the norm). It should be noted that reader experience is difficult to define and measure and that the interaction of “experience” with specific devices would be hard to substantiate.
What additional equipment or technical resources are needed in order to operate stress-loading MRI compared with standard MRI?
A KI affiliated with a weight-bearing MRI manufacturer indicated that weight-bearing MRI systems have similar installation (“sitting”) requirements to conventional MRI devices.
Commercially available axial-loading devices appear to have no additional requirements compared with the use of the same MRI devices in unloaded conditions; based on information available on the Web site of the manufacturers of the DynaWell axial compression system, the device can be used in “all known CT and MRI scanners on the market.”55
Dedicated extremity MRI devices appear to have significantly reduced technical requirements for installation and operation. Specifically, they have a smaller size (allowing them to be installed in relatively small spaces), do not require shielding of the room in which they are installed (because they include a small Faraday cage that provides shielding), and do not require a special power supply or air conditioning.
What is the cost of imaging with stress-loading MRI as compared with other imaging alternatives?
We attempted to collect information on the cost s associated with different types of MRI devices, particularly with extremity and upright MRIs as compared to conventional MRIs. KIs generally agreed that overall costs to health care facilities for obtaining and operating open, upright, and extremity-specific MRI scanners were lower compared with conventional MRI devices, as such devices have lower purchase and installation costs (e.g., costs for magnetic field shielding). KIs also added that, while health care facilities may be able to reduce costs by using this group of devices, the cost savings are typically not reflected in patients’ billing charges. We did not identify additional information on the cost of obtaining specific MRI devices.
Perusal of insurance company Web sites indicated that most policies assign two billing codes for MRI imaging, one for images generated by devices with low magnetic field strengths (< 1.0 T), and the other for images generated by devices with high magnetic field strength (> 1.0 T). We could find no separate billing code that differentiated between weight-bearing and non-weight-bearing imaging, or imaging obtained in different positions (e.g., flexion/extension). Based on input from our KIs, insurance companies typically reimburse patient s for one image per visit. All stakeholders confirmed this assertion. None of our stakeholders had knowledge of the exact charge (to the patient or payer) for obtaining an MRI image by billing code, or what the difference in cost between the two billing codes might be. A 2007 Technology Assessment of upright MRI commissioned by the Washington State Health Care Authority did report an estimated cost of $1,450 for a single image from an upright MRI, and costs for obtaining additional views ranging from $350 to $1,200 based on information obtained from manufacturers.24
Two additional issues concerning MRI imaging costs emerged during our interviews with stakeholders: (1) the potential for technically inadequate MRI images requiring a second exam with a conventional MRI, and (2) the potential for emerging MRI devices to generate multiple images in a single exam, leading to multiple billing and increased costs. The first issue concerned the marketing of emerging MRI devices with lower strength magnets by private clinics directly to consumers. Clinician KIs recounted examples where patients had decided to obtain upright MRI scans (with the implication that it was without their doctor’s recommendation) only to receive poor quality images that were insufficient for clinical decision making (particularly regarding surgical planning). These patients were often required to obtain additional MRI imaging in conventional scanners. The KIs noted that, in such cases, patients may be required to pay out-of-pocket for the second MRI examination. The issue of image quality may also be important given that several attorneys appear to advocate the use of images obtained by MRI devices under stress loading (particularly upright weight-bearing MRI) for evidentiary purposes.b
The second issue, put forth by the public payer KI, was that, as of the time of the interview, no specific billing code was available for the upright/positional MRI and that, due to the ability of the Fonar Corporation’s positional MRI to generate multiple “views” during one imaging session, each “view” was billed separately. The KI indicated that, on average, this has resulted in the number of scans being billed per patient visit at the positional MRI facilities in Washington State to be 2.5 times higher compared with that of conventional MRI facilities.
In summary, the emergence of new technologies and the diversification of MRI devices has the potential to further magnify the problem of cost and raise a new set of concerns regarding the relative quality and cost effectiveness of imaging using different types of devices, as well as who should bear the increased cost-burden.
Evidence Map
As described in the Description section, we focused on three features of emerging MRI technologies and applications to define seven non overlapping categories (Figure 3) of MRI devices or techniques. Using this classification scheme, we defined four categories of stress-loading MRI technologies of interest for the evidence map (Table 3). These categories serve as operational definitions that we employed for the purpose of this report, and do not necessarily imply that the specific MRI technique has been utilized in the published studies. Indeed, our literature searches did not identify any studies investigating devices that would fall under category D.
Table 3
Definitions and examples for the MRI device or technique groups considered.
Evidence Map of All Eligible Studies
Our MEDLINE search yielded 5,984 citations, 326 of which were retrieved in full text. Full-text articles were screened based on study eligibility criteria, yielding 55 publications that used MRI with weight-bearing or stress-loading protocols in patients with musculoskeletal conditions.25–28,56–106 Of these, one paper reported data from two separate studies.26 One additional paper was identified through hand searching of reference lists.107 Thus a total of 57 studies (in 56 publications) were included in our evidence map.
We categorized these 57 studies according to our definitions for emerging MRI technologies under weight-bearing or loading stress as specified in Figure 3(areas A to D). Based on these definitions, 36 studies fell under category A (i.e., open, positional, and weight-bearing MRI),25–28,56–84,101,106,107 two studies fell under category B (i.e., use of specialized devices to obtain MRI imaging under weight-bearing conditions in a closed MRI scanner),86,103 and 19 studies fell under category C (i.e., use of specialized devices to “simulate” gravity [i.e., axial loading] in a conventional MRI).26,85,87–100,102,104,105 None of the qualifying studies fell under category D.
It should be noted that multiple studies originated from the same research centers, and it is often not possible to ascertain whether patients or controls were shared between studies conducted by the same investigators. Patient population overlap creates the impression that more studies are available on a given clinical question than may be the case. In an effort to explore whether particular findings or research groups were over-represented in the literature, we generated a graph to depict groups of studies that had overlapping author lists (Appendix D). Twenty publications (36 percent of those reporting on eligible studies) were produced by four teams. One team that has published 6 manuscripts, corresponding to approximately 10 percent of all studies, includes the inventors of the DynaWell axial loading, Drs. Danielson and Willen, as co-authors.55
Below, we present a summary of all 57 studies followed by a more detailed presentation of the characteristics of studies falling under each device category. Although these analyses include the comparative studies we identified (i.e., studies that applied at least two diagnostic tests on the same patient population and investigated clinical outcomes), these studies are also further discussed below in Comparative Studies That Reported Clinical Diagnostic or Patient Outcomes, and the characteristics of these studies and the outcomes they assessed are presented in Appendix E.
Characteristics of Eligible Studies
All of the studies were published between 1993 and 2010. The most commonly imaged body regions were the lumbar spine (33 studies) and knee (13 studies). Figure 4 presents the eligible studies stratified by study design and anatomic region assessed. Across all studies, the median of the mean/median age of patients with musculoskeletal diseases or conditions was 42.6 years (25th–75th percentile: 31.6, 50); the median mean/median age of controls (for case-control studies) was substantially lower at 29.9 years (25th–75th percentile: 28–34.4). Approximately 50 percent of the individuals included in the eligible studies were male (equally distributed in both patients and controls).
In general, studies were small; the median number of included cases was 26 (25th–75th percentile: 17, 45) and the median number of controls was 13 (25th–75th percentile: 12, 20; for case-control studies only). No randomized controlled study or nonrandomized comparative study of testing versus no testing was identified. The majority of studies were cross-sectional (37 studies), or had case-control designs (13 studies). Only five longitudinal studies and two studies obtaining imaging pre- and post-application (within minutes) of an orthopedic intervention or physical activity were included. The vast majority of studies did not systematically identify cases or controls (convenience sampling). Fifteen studies (27 percent) were comparative studies of two diagnostic tests and reported on clinical outcomes. Figure 5 presents the eligible studies stratified by study design and weight-bearing or stress-loading device used.
Patient-relevant outcomes were assessed infrequently (five studies). Most studies (27 studies, 47 percent) focused on the feasibility (defined as agreement in anatomic measurements between different imaging modalities as the outcome of interest) of imaging under weight-bearing or stress-loading conditions. Most studies (45 studies) exclusively enrolled symptomatic patients, 7 studies enrolled exclusively a symptomatic patients and 4 studies enrolled mixed populations (one study did not report this information). Only 14 studies assessed accuracy outcomes (we defined this broadly as a diagnosis of abnormality in symptomatic patients made based on weight-bearing or stress-loading MRI), 1 study addressed patient management or treatment planning, and 2 studies reported on disease monitoring.
Only 10 studies reported harms or adverse events associated with weight-bearing MRI testing (Table 4). In the studies that reported relevant information, most adverse events were new-onset or worsening pain/neuropathy while the patients were placed under loading stress (weight-bearing or axial loading). Studies reported test interruption and incompletion rates of 5 to 10 percent due to symptoms developing during stress loading; and one study reported amending its design (evaluating sitting instead of upright MRI) because patients could not stand still during the upright exam.74
Table 4
Studies of weight-bearing or stress-loading MRI that reported information regarding adverse events.
Most studies were conducted outside the United States (see also Appendix D). Funding information was often not reported; among studies that reported relevant information, frequently no funding source was identified (“no funding was received”). Given the lack of relevant information in a large number of studies and the potential influence of different editorial policies on reporting financial support, it is difficult to interpret this finding.
Provided below is a qualitative summary of the findings regarding the populations studied, outcomes assessed, and reporting completeness in the eligible studies, arranged by categories of MRI device. Table 5 summarizes the characteristics of the patient populations or demographics, and Table 6 summarizes the study design characteristics of eligible studies.
Table 5
Summaries of descriptive characteristics of the studies considered eligible—patient characteristics or demographics,.
Table 6
Summaries of descriptive characteristics of the studies considered eligible—study design characteristics,.
Open, Positional, and Weight-Bearing MRI
Thirty six studies were classified as “open, positional, and weight bearing MRI” systems (category A). In general, studies were small; on average they included 101 cases (median = 30; 25th–75th percentile: 20, 50) and 20 controls (median = 13; 25th–75th percentile: 12, 20; for case-control studies only). The majority of studies were cross-sectional (24 studies), or had case-control designs (7 studies). Only four longitudinal studies were identified. Followup duration was less than a year (reported in three of the four studies). One study had a pre-post design, in which weight-bearing MRI was used to assess outcomes of spinal manipulation interventions (imaging was performed before and immediately after the intervention). The most commonly imaged body regions were the lumbar spine (20 studies) and knee (6 studies). Clinical outcomes were assessed infrequently; the majority of studies (25 studies) reported on anatomic measurements or rater agreement under weight-bearing or stress-loading conditions, 9 studies reported on accuracy outcomes, 2 studies reported on impact on diagnostic thinking, and no study reported on patient-centered outcomes.
MRI Imaging Under Weight-Bearing Conditions in a Closed MRI Scanner
Only two studies (both with the same first author) assessed MRI imaging under weight-bearing conditions in a closed MRI scanner (category B).86,103 The investigators used a positioning device with a section cut out to permit uninhibited movement of the patellofemoral joints in a conventional MRI (prone position) for capturing dynamic (kinematic) images under loaded or unloaded conditions.
The first study was conducted in 1993 among 19 patients with a clinical diagnosis of abnormal patellar alignment and tracking. The authors reported that the positioning device “will soon be commercially available for use with Signa MR imaging systems (GE Medical Systems) and can be easily modified for use with other MRI systems.” The main findings of this study are described later in this report in Comparative Studies That Reported Clinical Diagnostic or Patient Outcomes.
Subsequently, the same positioning device and kinematic MRI protocol were used to evaluate the effect of applying a stabilizing brace to 15 patients who had hallmark signs and symptoms of patellar malalignment. This study was published in 2000 and reported that the positioning device had become commercially available.
“Simulated” Gravity in Conventional MRI Scanners
Nineteen studies used an axial-loading device to “simulate” gravity in conventional MRI scanners (category C). In general, studies were small; on average they included 40 cases (median = 24; 25th–75th percentile: 12, 34) and 20 controls (median = 14; 25th–75th percentile: 13, 18; for case-control studies only). The majority of studies were cross-sectional (13 studies), or had case-control designs (5 studies). One study had a longitudinal design but did not clearly report the duration of followup. The most commonly imaged body regions were the lumbar spine (13 studies) and the knee (5 studies). It was infrequent that clinical outcomes were assessed; most studies focused on the feasibility of imaging under weight-bearing or stress-loading conditions (9 studies); 13 studies reported on anatomic measurements/rater agreement, 4 studies reported on diagnostic accuracy, one study reported on the impact of the tests on diagnostic thinking, one study reported on the impact on treatment decisions, and no study reported on patient-centered outcomes.
In summary, our evidence map showed that studies of stress-loading MRI are small in sample size and employ study designs of relatively low internal validity. Outcomes are also not immediately clinically applicable, with many studies (approximately 50 percent) focusing on the feasibility of imaging under weight-bearing or stress-loading conditions (for example, agreement in anatomic measurements between different imaging modalities was a common outcome). Very few studies (12 percent) reported on clinically relevant outcomes. More details of these studies are described in following section.
Comparative Studies That Reported Clinical Diagnostic or Patient Outcomes
The most direct applicable study designs for clinical decisionmaking are studies that compare two or more diagnostic strategies, follow the patients through decision and treatment, and then report on patient outcomes. However, none of the comparative studies were in such design.
Of the 57 studies discussed above, 15 compared two diagnostic tests and reported either clinical diagnostic or patient outcomes.27,58,70,71,73,82,84–87,89,96,98,104,105 Of these, four compared open, positional, weight-bearing MRI in lumbar spine imaging with four different comparative tests27,70,82,84; seven compared axially-loaded images of the lumbar spine with preloaded images in the same conventional MRI scanner;85,87,89,96,98,104,105 and four compared weight -bearing or stress-loading MRI with MRI without loading for diagnosis of extremity abnormalities.71,73,86 None of these studies used a “gold standard” (e.g., surgical findings) for their diagnoses. Only four studies reported patient outcomes (pain, anxiety, testing preference, or physical function),82,87,89,98 and only one study reported changes in patient management based on additional information gained from axial loading MRI.104 (Appendix E)
Of the four studies of lumbar spine imaging comparing open, positional, weight-bearing MRI, two small studies (enrolling ≤50 patients) reported that open, positional, weight-bearing MRI contributed additional information to diagnoses compared with conventional (non weight-bearing) MRI. However these studies did not report impact on treatment choice or patient outcomes.70,82,84 Another study assessed patient preferences and anxiety during open, positional MRI and during lumbar myelography in 30 subjects, and reported that more patients were anxious during myelography than during MRI and that more patients preferred MRI than myelography.82 However there was no data on whether diagnosis, treatment, or patient outcomes were affected.
The fourth study compared diagnosis of lumbar abnormalities based on upright MRI in the extension or flexion positions to diagnosis based on upright MRI in the neutral position.27 Subjects were 533 patients with different grades of disc herniation. All positions (extension, flexion, or neutral) were performed while patients were standing (thus all were weight-bearing positions). A range of “missed diagnoses” by upright MRI in the neutral position as compared to upright MRI in the extension or flexion positions was reported. It should be noted that the reported “missed diagnoses” assumed upright MRI was the reference standard and no functional outcomes were reported.
All seven studies of lumbar spine imaging comparing axially loaded images with preloaded images in the same conventional MRI scanner reported that use of axially loaded MRI led to additional diagnoses or had impact on diagnostic thinking.85,87,89,96,98,104,105 One study also reported good surgical outcomes among patients whose hidden stenosis was disclosed by axial-loading MRI, but there was no control group for comparison.98 However, it should be noted that five of the seven studies came from the same group of investigators in Sweden, and that there were obvious overlaps in patients reported in the five studies.87,96,98,104,105(see also Appendix D). At least two of the investigators in this group are the coinventors of an axial-loading device, DynaWell L-Spine, which is currently commercially available. All seven studies suffered from potential selection and/or verification biases.
Two of the four studies comparing weight-bearing MRI with MRI in the supine position (not weight bearing) found that the two techniques were comparable and that weight-bearing MRI did not provide additional information for the diagnosis of plantar fasciitis or Morton’s neuroma.71,73 Another study included only patients who had a prior diagnosis of meniscal tears by conventional MRI and confirmed by arthroscopy; weight-bearing MRI was not used to provide additional information for the diagnoses. This study reported that patients with displaceable meniscaltears (diagnosed by weight-bearing MRI) had significantly more pain than patient with nondisplaceable meniscal tears.58 The last, industry-funded, case-control study reported that loaded dynamic MRI produced significantly less missed diagnoses of patellofemoral joint abnormalities than did unloaded dynamic MRI.86 However, no functional outcomes were reported to verify the importance of the imaging findings.
Ongoing Studies in ClinicalTrials.gov
Our search for ongoing clinical trials utilizing weight-bearing or stress-loading MRI identified three ongoing studies: NCT00665548, NCT00706459, and NCT00887744.
Briefly, the first study is a collaborative case-control study currently being conducted by the University of California, San Francisco and Pfizer. The study aims to enroll a total of 33 female subjects older than 40 years of age. Cases will be osteoarthritis patients, while controls will be healthy volunteers. The goal of the study is to compare two modalities (x ray and MRI) for imaging the knee joint under both weight-bearing and non-weight-bearing conditions. A 3.0T scanner will be used to obtain all MRI scans. The study’s record on ClinicalTrials.gov indicates that data collection for the primary outcome measure was completed in February 2009, but we could not identify any related publication.
The second study is also being conducted by investigators at the University of California, San Francisco, with funding from the National Institutes of Health. Also a case-control study, the total enrollment target is 105 subjects of both sexes, between 25 and 60 years old. Cases will be patients with lumbar back pain scheduled for back surgery, patients with degenerative disease without classic discogenic back pain, and patients who have undergone discectomy for herniated discs. Controls will be age-matched volunteers without back pain. All study subjects will undergo lumbar spine imaging using a 3.0T MRI scanner (including axially loaded scanning) to assess whether this method can be used to identify painful degenerated discs in patients with chronic back pain. The investigators indicated that this study would serve as a pilot for a larger trial.
The third study is an interventional, multicenter, single arm, post-marketing, clinical followup study. The primary purpose of the study is to measure the change in severity of symptoms and ability to function in everyday activities in patients suffering from degenerative lumbar spinal stenosis after treatment using the Aperius PercLID device. The expected enrollment for this study is 163 subjects of both sexes older than age 21. A secondary outcome of the study involves positional MRI scanning to measure the changes in spinal canal, foramina, and disc, immediately post-operatively and after 12 months of followup.
Enrollment for all three studies was completed in 2009; however, their results are not yet available on the ClinicalTrials.gov Web site, and we could not identify any corresponding publications in MEDLINE. Thus we are not certain the motivation or purpose of including the stress-loading MRI in these studies. It should be noted that these three studies are small in sample size and conducted in three different patient populations, so it is unlikely that their results, once published, would change the conclusion of our evidence map.
Projected Uptake and Potential Growth
All KIs suggested that MRI under weight-bearing or stress-loading conditions is an actively growing research field. Several KIs indicated that this should be a key direction for future research for radiology in general. However, the majority of the studies conducted to date appear to be observational in design based on convenience samples of patients and healthy controls. Such studies are often not registered on ClinicalTrials.gov, and their results may not be generalizable to clinical settings. Industry sources also suggested that future developments in MRI equipment are likely to focus on imaging specific joints or organ systems in physiologic conditions, instead of the current practice of using whole-body, conventional MRI systems for all diagnostic purposes. The findings of our literature searches indicate that most studies of stress-loading MRI conducted to date pertain to the lumbar spine and the knee joint applications. Our industry KIs did not disclose specific projections for future uptake of their respective technologies; however, based on information from Fonar Corporation’s official Web site, a new open, in-office (small-bore), multipositional extremity MRI (mpExtremity MRI) that allows weight-bearing imaging of lower extremities in a standing position is being developed. 108 Our findings on dedicated extremity MRI scanners (summarized in Appendix F) indicate that this is a more mature imaging technology and suggest that a systematic review of the available literature may be feasible.
Clinically oriented KIs suggested that a stepwise approach to the further development of weight-bearing or stress-loading imaging would be preferable. Initially, studies would be conducted to standardize diagnostic methods across centers and ensure that images of adequate quality could be obtained. Specifically, clinicians contended that research efforts ought to be bent toward developing imaging protocols that mimic orthopedic clinical examination, such as imaging in flexion/extension or under stress loading. After these initial steps, larger validation studies should be undertaken, followed by appropriately controlled studies to assess the impact of using weight-bearing MRI on clinical or patient outcomes. We expand on these suggestions based on a proposed analytic framework in the Summary and Implications section of this report.
Regarding the selection of outcomes for future studies, KIs agreed that diagnostic tests should be judged by the amount of additional information they offer as compared to other imaging methods. Clinicians and stakeholders suggested that studies should look beyond diagnostic accuracy and investigate the impact on diagnostic decisionmaking, clinical treatment decisions, and patient outcomes. Additionally, both clinician and payer KIs mentioned costs as an important aspect of diagnostic decision making and suggested that the cost-effectiveness of MRI technologies should be established by further research. Industry KIs suggested that the cost-effectiveness of their respective products was evident.
Footnotes
- a
As discussed in the Methods section, a brief discussion of dedicated extremity MRI is included in Appendix F, due to the technology’s potential for rapid development and diffusion.
- b
This issue was identified by one of the peer reviewers of this Technical Brief and is easily verifiable by simple Internet searches.
- Findings - Emerging MRI Technologies for Imaging Musculoskeletal Disorders Under...Findings - Emerging MRI Technologies for Imaging Musculoskeletal Disorders Under Loading Stress
- LP_RS09695 [Lactobacillus plantarum WCFS1]LP_RS09695 [Lactobacillus plantarum WCFS1]Gene ID:1062579Gene
- KCMF1 potassium channel modulatory factor 1 [Bos taurus]KCMF1 potassium channel modulatory factor 1 [Bos taurus]Gene ID:613522Gene
Your browsing activity is empty.
Activity recording is turned off.
See more...