Research News: 2002 - 2003 - 2004

Morphological and Functional Musculoskeletal Imaging
Author: Sharmila Majumdar, PhD

With the aging population of the United States, degenerative diseases affecting the joints, spine and other skeletal sites are growing to be a major source of morbidity, declining quality of life, and are taking a serious financial toll on society. Imaging has made a tremendous impact in diagnostic procedures, in surgical planning, and guided surgical applications. However, beyond anatomical and subjective depictions of anatomy, quantitative, morphological and functional musculoskeletal imaging methods are still underutilized. Over the last two years, there has been a focused attempt to increase collaborative efforts and address the development of quantitative musculoskeletal imaging in the Department of Radiology at the University of California, San Francisco. Spurred on by a Bioengineering Research Partnership Grant from the National Institutes of Aging, participants from UCSF (Sharmila Majumdar, PhD, Lynne Steinbach, MD, and Cynthia Chin, MD, in Radiology; Jeffrey Lotz, PhD, and Michael Ries, MD, in Orthopedic Surgery; Karen King, PhD, in Medicine), Lawrence Berkeley National Laboratories (Thomas Budinger, MD, PhD) and industrial partners have focused on the systematic study of the morphology and function of the musculoskeletal system in disease and health.
The team effort utilizes the strength of each individual partner at the inception and throughout the design of technological advances. This sets a format for interaction, which prevents the isolated development of tools and techniques that are irrelevant and unrealistic in the clinic, or need elaborate modifications and redesign at the time of clinical utilization and commercialization. In the development of these imaging techniques, the inclusion of corporate partners establishes the pathway for utilizing already established prototypes, use of corporate resources, rapid standardization, testing and clinical dissemination of the techniques. While the long-term objective of this consortium is to understand the link between morphology, function and clinical symptoms in the musculoskeletal system, an immediate objective has been to develop, implement and optimize novel non-invasive imaging methods (magnetic resonance imaging [MRI] and positron emission tomography [PET], infra-red imaging, and micro- computed tomography) that will allow us to depict the musculoskeletal system, quantitate morphology, function, provide unique 3-D visualization and graphic representations of function and morphology. In particular, the goal has been to leverage the quantitative aspects of imaging to further complement the characterization of degenerative knee, and the spine and enhance the clinical utility of imaging.

Disorders of the Spine
Disorders of the spine have a tremendous impact on society, both physically, through the morbidity of afflicted individuals, and financially, through lost productivity and increased health care costs. Back pain is the second leading cause for ambulatory care in the United States and direct medical costs for this condition are estimated at over $20 billion per year. Medical back problems' comprised the second most common medical diagnosis-related group for all hospital discharges in 1987, following only normal childbirth. Despite the significance of this problem, the etiology of symptoms is diverse and unclear in many patients, and consequently there are few reliable methods by which to prospectively determine the appropriate course of patient care (i.e. conservative management or surgery).

Three principal causes for patient complaints are: abnormal motion-instability; tissue inflammation; and foraminal stenosis or narrowing of the foraminal space, which may result in nerve compression. With these factors in mind, the last two years have led to the development and research for the optimal imaging modality which would contain quantitative information regarding these factors above to best aid in establishing the patient's diagnosis. Optimizing the use of novel MR imaging, such as line scan diffusion techniques, Chin, David Newitt, MD, and Bashir Taoli, MD, are exploring the possibility of improved quantitative characterization of spinal conditions. As seen in Figure 1, a fat suppressed image in a patient with a plasma cell tumor shows a bright region, and a corresponding diffusion image to the right shows a significant dark tumor region. Images such as these, that reflect differences in tissue water diffusion, a characteristic of tissue composition and biochemistry, provide a means of quantitatively describing pathological changes.

Computed tomography methods, using newly installed multi-detector scanners exploring age-related narrowing of the spinal canal and narrowing of the foraminal space, are also under investigation. In Figure 2,Thomas Lang, PhD, has generated a three-dimensional rendering of a vertebral body from a stack of computed tomography images. The figure shows some of the metrics that can be measured using such images. Using these metrics, we have found that subjects who had reported symptoms such as pain, numbness, and weakness in the legs, tended to have lower bone density but larger vertebral size than normal subjects. Consistent with expectations, the spinal canal area tends to be smaller in subjects with symptoms, and the ratio of vertebral cross- sectional area to spinal canal area is larger. There was a tendency for symptomatic subjects to have more osteophytes and more narrowing of the lateral recess than the normal subjects, as well as potentially a trend for symptomatic subjects to have more narrowing of the foramina.

In collaboration with UC Berkeley's Jitendra Malik, PhD, and Sergie Belonghi, PhD, Bioengineering graduate student Julio Carballido has developed methods for segmenting the vertebral bodies as seen in Figure 3. This combination of state-of-the-art MR methods with CT imaging and computerized image analysis holds tremendous potential for studying lower back pain, degeneration of the spine, and age-related spinal stenosis. Proposed studies using PET (Budinger) will further enhance the research in this area, bringing molecular imaging into the realm of the clinic in the foreseeable future.

Degenerative Diseases of the Knee Joint
Joint pain and physical impairment are responsible for extensive use of medical and surgical resources in the United States. The important tissues of the joint are the bones, the synovial membrane, the cartilage on the surfaces of the articulating bones, which provide a low coefficient of friction surface allowing smooth joint motion, and the ligaments and tendons, which attach the articulating surfaces or are attached to these surfaces. Osteoarthritis (OA) and related disorders accounted for 85% of all total knee replacements and total costs for knee and hip replacements are well over $300 million.
The course of OA is challenging to describe. It is a heterogeneous and multifactorial disease characterized by the progressive loss of cartilage and the development of altered joint alignment, as well as changes in the adjoining bone. The integrity of the articular cartilage and the inter- relationship between the bone and adjoining cartilage is very important in maintaining joint stability and preventing joint degeneration and an impairment of function. Physicians seek safe and effective ways to treat OA and musculoskeletal disability. To assess therapy, diagnostic tools that quantitatively measure progression, degeneration and yet correlate with clinical measures, are essential. A better understanding of the etiology, biochemical changes, metabolic changes and kinematics or motion studies of the degenerative joint should be the first step in attempting to design effective surgical interventions and therapeutic regimens to promote repair after injury and prevent joint failure and degeneration.
With these goals in mind, high-resolution magnetic resonance images have been developed to image the cartilage as well as the bone in the knee joint. As seen in Figure 4, the high signal cartilage can be defined, and its volume and thickness not only can be visualized, as seen in the figure, but also quantified. The relationship between bone changes and cartilage changes in osteoarthritis remains a major question, and unraveling the timing of these changes will have a major impact on understanding the pathophysiology of OA as well as devising treatments. Using the images depicting the fine bone network as seen in Figure 4, the link between such cartilage-bone interactions may be unearthed.

OA is not a disease of static joints, but one that is mediated by biomechanical loading as in gait, and understanding the alignment of the bones and meniscus is of some importance in assessing the status of the joint. Previous studies of knee motion have been limited by invasiveness, two- dimensionality, accuracy of marker positioning, or lack of physiologic weight bearing. Using a novel device developed by Vikas Patel, MD, an Orthopedic Surgery resident, three-dimensional, MR studies that replicated load bearing in the magnet were undertaken. Tibio-femoral and patello- femoral motion was studied, first in normals and then in subjects with instabilities and osteoarthritis. Images were obtained in different positions of flexion and extension, as shown in Figure 4. The relative motion of the bones (patella and femur in Figure 5), the translation, and the area of contact, was quantified.

MR images of cartilage also provide quantitative information pertaining to relaxation times such as T2, which depend on the collagen content and chemical composition of articular cartilage. A Bioengineering graduate student in Majumdar's laboratory, Srinka Ghosh, worked in close collaboration with Michael Ries, MD, of Orthopedic Surgery and obtained specimens of the cartilage from subjects undergoing total knee replacement. Longer T2 values are accompanied by a histological confirmation of disorientation of the cartilage fibers, fibrillation and fissures. Following up on this tissue characterization work, Tim Dunn, a second year Bioengineering graduate student, has developed a method for mapping T2 changes in human subjects. The T2 values obtained in the cartilage are expressed as a ratio of the T2 values in healthy subjects, overlaid as a color map on the MR images (Figure 6). The shades of blue and purple are close to a value of 0, and correspond to the healthy range of T2 values; yellow and red reflect higher values corresponding to subjects with severe osteoarthritis.

In an effort to complement the imaging studies, King, in the Dept. of Medicine, has been collaborating with investigators in Radiology, and analyzing serum samples from the subjects who have undergone MR scanning. She has used ELISA analysis to quantify molecular markers for joint degeneration (Cartilage oligomeric matrix protein [COMP] and Serum CILP, a cartilage-specific protein, also from the extracellular matrix). Initial results demonstrate that as total cartilage volume decreases, serum COMP and CILP increase. These data also show that a moderate correlation exists between a subject's serum COMP level, as measured by ELISA, and the amount of cartilage in medial tibia, as measured by MRI.
The investment of $4.5 million made by the National Institutes of Aging for the Bioengineering Research Partnership for musculoskeletal imaging involves combining a number of state-of-the-art methods to characterize the joint, and spine develop imaging, as well as serum biomarkers. This consortium of specialists, collaborating across disciplines, are working toward developing techniques that ultimately will be used to assess the clinical status of the subject more effectively.