BBI Contributing Editor

ARLINGTON, Virginia The 35th annual Oak Ridge Conference, held here in April, focused on new approaches to diagnosing neurological diseases, and covered the latest developments in laboratory testing for such disorders as well as in vivo imaging and monitoring technologies that are proving valuable in patient management. While attendance at the conference sponsored by the American Association for Clinical Chemistry (AACC; Washington) was quite low, reflecting both a general trend in the device and diagnostics sectors and the niche focus of the meeting, a number of technologies were described that may form the basis for a new high-growth segment of the diagnostics market.

Neurological disorders, including mental or behavioral disorders, are among the most common illnesses worldwide. Major types of neurological disorders include stroke, cerebrovascular aneurysms, central nervous system tumors, arteriovenous malformations, hydrocephalus and Parkinson's disease, as well as mental or behavioral disorders such as Alzheimer's disease, epilepsy, depression, schizophrenia and mental retardation.

The World Health Organization (Geneva, Switzerland) says that one in every four people develop one or more mental or behavioral disorders at some stage in life. Five of the 10 top diseases causing disability worldwide are psychiatric disorders, and major depression ranks No. 1 as a cause of disability (see Table 1 on page 146). The major types of neurological disease afflict more than 40 million persons in the U.S. alone.

The diagnosis of neurological diseases is often challenging. Very few good laboratory tests have been developed so far, so diagnosis typically relies on psychological exams in the case of behavioral disorders such as Alzheimer's disease, schizophrenia or depression. Non-invasive imaging modalities, particularly magnetic resonance imaging (MRI), are becoming increasingly important tools for the diagnosis of stroke, aneurysms and a number of other neurological disorders, but using such methods as initial screening tests to identify at-risk individuals is impractical and lacking in sensitivity.

Diagnosis of Alzheimer's disease typifies many of the challenges that have been encountered in the development of tests for neurological disorders. There is a need for tests that can pinpoint the specific disease process other than brain biopsy, which can only be performed post-mortem. As discussed at the Oak Ridge gathering by John Hardy, PhD, of the National Institute on Aging (Bethesda, Maryland), the genetic basis of Alzheimer's disease is beginning to be unraveled, setting the stage for the development of tests that can be used in clinical practice to improve diagnosis. Presence of the apolipoprotein e4 allele is one well-known risk factor. In addition, three genes have been identified, for amyloid precursor protein, presenilin-1 and presenilin-2, that when mutated can result in abnormal accumulation of amyloid plaques and tau proteins.

Mutations in the presenilin genes are now known to be responsible for 80% to 90% of early-onset Alzheimer's disease. It is not yet clear if late-onset Alzheimer's is caused by the same mutations, but Hardy said he believes that will prove to be the case. At present, there are only a few labs offering genetic testing for mutations related to Alzheimer's. Among the commercial providers are Athena Diagnostics (Worcester, Massachusetts), Kimball Genetics (Denver, Colorado), Spectrum Health (Grand Rapids, Michigan) and HUCH Laboratory Diagnostics (Helsinki, Finland).

While genetic testing has now become a major segment of the clinical laboratory services market, genetic tests for neurological disorders comprise only a small percentage of the total. As shown in Table 2, below, the U.S. market for DNA-based genetic testing services is estimated at about $224 million, and is expected to grow at 15% to 17% annually over the next few years. There are 567 laboratories worldwide performing genetic tests for a total of 948 diseases, according to the database.

While testing for Alzheimer's disease remains a relatively small niche, there has been significant progress in developing new tests for multiple sclerosis (MS), including identification of candidate serum markers as well as refinement of in vivo imaging tests. While the prevalence of MS is only about one-10th that of Alzheimer's, there are an estimated 2.5 million people afflicted with the disorder worldwide. Furthermore, there have been some recent advances in development of drugs to treat MS. As discussed by Henry McFarland, MD, of the Institute of Neurological Disorders and Stroke (Bethesda, Maryland), the hallmark of the disease is the loss of the myelin sheath that surrounds neurons, and a resultant slowing of the transmission of neural signals.

A breakdown of the blood-brain barrier precedes the loss of myelin, followed by axonal loss. Brain N-acetylaspartate (NAA) is used as a marker of axonal loss, and MRI is used to track cerebral atrophy as the disease progresses. Another marker is soluble VCAM, a cell adhesion molecule. Interferon treatment targets changes in cell adhesion molecules associated with MS, and those changes can be tracked by monitoring sVCAM. Most recently, researchers have begun using proteomics technology to discover new markers for MS. Microarray analysis of blood samples from MS patients who are responsive and non-responsive to drug therapy has identified two interleukin markers, IL-8 and IL-12, that show good correlation with response to therapy. An IL-8 assay is available outside the U.S. on the ImmuLite and ImmuLite 1000 immunoassay analyzers from Diagnostic Products Corp. (Los Angeles, California).

Proteomic analysis has been used in the discovery of new markers for a number of other neurological diseases. For example, new markers have now been identified for cerebral palsy and brain tumors using proteomic chips as well as mass spectrometry. At the Oak Ridge conference, Ebenezer Satyaraj, PhD, of Molecular Staging (New Haven, Connecticut), described development of a protein chip used to analyze cytokine expression in cerebral palsy patients. Each chip contains 16 sites on a glass substrate, with each site having different anti-cytokine capture antibodies immobilized on the surface. A sandwich immunoassay format is used along with Molecular Staging's Rolling Circle Amplification to provide a two- to three-log amplification of signal. In the study of cerebral palsy markers, multiple chips are used to analyze the expression of 78 cytokines in cord blood samples taken at birth, with a goal of identifying a set of markers that can be used to detect the disease at an early stage. About 10,000 babies are born each year in the U.S. with cerebral palsy, with a resultant cost of $5 billion to the U.S. healthcare system.

Molecular Staging has analyzed cytokine expression patterns in cord blood to identify a classifier set of 15 cytokines and related molecules, including markers such as epidermal growth factor, ciliary neurotrophic factor, IL-12, IL-13 and IL-15, that provide a cross-validation rate of 94% for detection of cerebral palsy. A 42-marker panel gives a predictive value of 98%. The markers also may serve as potential therapeutic intervention targets. In principle, early diagnosis could allow treatment to be initiated at birth, and mitigate much of the irreversible neurological damage that occurs.

The same marker panel may have additional applications in the diagnosis of inflammatory bowel disease and Molecular Staging also has developed protein chips for ob/gyn, cancer and organ damage testing. Dominic Desiderio, PhD, of the University of Tennessee Health Science Center (Memphis, Tennessee), described the use of 2-D gel electrophoresis coupled with MALDI-TOF mass spectrometry to analyze protein expression in patients with pituitary adenomas, with a goal of identifying markers useful in diagnosing pituitary cancers.

Studies also have been conducted using cerebrospinal fluid in an effort to identify a panel of protein markers that are associated with idiopathic back pain. In the latter studies, expression of markers including proenkephalin A and Substance P has been found to correlate with response to drug therapy, potentially providing an objective method to assess patients with back pain, as opposed to the subjective methods used now that rely on patient self-assessment.

Major opportunity for stroke, stroke risk tests

Tests for the diagnosis of stroke, and ultimately for assessment of the risk of stroke, represent one of the largest potential market opportunities within the neurological testing segment. There are currently no rapid in vitro diagnostic tests for stroke, as there are for myocardial infarction, even though the annual incidence of stroke in the U.S. is something more than half as large as the estimated 1.1 million cases of new or recurrent myocardial infarction. In addition, there are no specific markers that indicate a predisposition to stroke, although there are preventive therapies that can be implemented in patients who are found to be at risk based on symptoms such as transient ischemic attack and imaging studies.

Ken Buechler, PhD, of Biosite Diagnostics (San Diego, California), who discussed the development of new stroke markers, said there are 788,000 emergency room visits in the U.S. each year related to stroke or suspected stroke and an estimated 167,661 deaths annually based on 2003 statistics from the American Heart Association (Dallas, Texas). The cost to the U.S. healthcare system for stroke is estimated at $51.2 billion. In part because of the lack of effective rapid diagnostic methods for stroke, the number of stroke patients who receive treatment within the three-hour window needed to achieve benefit is a small percentage of the total. Less than 4% of stroke patients receive thrombolytic therapy, according to Buechler. There are needs for tests that can allow physicians to rule in or rule out stroke in the emergency setting, as well as for tests that can help determine the best care path for a stroke patient, such as tests to help select patients for thrombolysis and for prognostic tests.

Biosite, a supplier of point-of-care tests for myocardial infarction and drugs of abuse, is developing a panel of markers for stroke that has shown promise for rapid diagnosis. Such a test could reduce the time needed to decide to send patients for a CT scan, increase the number of patients who receive thrombolysis and reduce the length of stay in the emergency department for patients who are found to not have a stroke. In collaboration with physicians at Duke University (Durham, North Carolina), the company's Biosite Discovery unit has analyzed blood samples from stroke patients and identified five markers. They include S-100b (a glial factor), B-type neurotrophic growth factor, Von Willebrand factor (a coagulation factor), matrix metalloprotease-9 (MMP-9, an acute phase protein) and monocyte chemotactic protein-1 (MCP-1, an inflammatory factor) that together provide high sensitivity and specificity for stroke and may also provide prognostic information.

The panel does not distinguish between ischemic and hemorrhagic stroke, but Biosite does not view that as a drawback since hemorrhagic stroke can readily be detected with a computed tomography (CT) scan. The panel also does not detect transient ischemic attacks (TIAs), a desirable feature for a test intended to detect a true stroke.

Use of a panel of markers has proven to be a key factor in allowing development of a test with good sensitivity and specificity, since the markers exhibit rather poor performance when used alone, but each contributes significant improvements when added to the panel. Adjustment of threshold levels for each marker can allow the test parameters to be tailored to meet physician requirements. For example, 100% sensitivity for ischemic stroke can be achieved at a specificity of 93% for up to 48 hours following an event. Optimization of the marker response factors resulted in a final sensitivity of 96.2% at 97.2% specificity for samples taken within 24 hours of onset. That performance compares very favorably with CT scans, which exhibit a high specificity but a sensitivity of only 33% for ischemic stroke.

Another potential application is prediction of vasospasm following subarachnoid hemorrhage, one of the major forms of hemorrhagic stroke. Vasospasm occurs in 70% of cases of subarachnoid hemorrhage and is a leading cause of morbidity and mortality. Biosite Discovery has found that VWF, MMP-8 and vascular endothelial growth factor (VEGF) are all elevated significantly in patients who subsequently have vasospasm. Studies also have shown that the five-marker panel can differentiate damage related to head trauma from stroke damage. Use of rapid diagnostic tests for stroke could help to reduce inappropriate care, avoid the adverse impact on outcome of slow diagnosis of stroke, and lower total costs associated with stroke management. The Biosite test can be performed in 15 minutes using a small cartridge that performs blood filtration and lysis, incubation with antibody, signal generation and waste disposal. The detection zone consists of a protein array that can in theory measure up to 100 different analytes with a 30-second to one-minute incubation.

While there are clear benefits for tests that can provide a more sensitive and specific stroke diagnosis, an even more important need exists for tests able to determine stroke risk or predict stroke. CIS Biotech (Atlanta, Georgia), working in collaboration with a group of researchers led by Svetlana Dambinova, DSc, PhD, at Pavlov's State Medical University (St. Petersburg, Russia), is developing a blood test that detects autoantibodies to the NR2A/2B subunits of N-methyl-D-aspartate (NMDA) neuroreceptors. The expression of NMDA receptors is activated by release of glutamide and aspartate under conditions of neurotoxicity resulting from reduced cerebral blood flow.

Dambinova has found that autoantibodies are generated to NMDA receptors that can be detected in blood. The degree of reduction in blood flow due to a TIA is sufficient to cause autoantibody production, allowing identification of patients who are experiencing such attacks, which often precede an ischemic stroke. If a stroke subsequently occurs, NMDA autoantibody subunits are generated within three hours. No autoantibody subunits are generated in the case of hemorrhagic stroke.

The test can be used for the detection of ischemic stroke, as well as for the differentiation of ischemic vs. hemorrhagic stroke, and potentially can be used as a tool to assess stroke risk. A 10-minute assay has been developed for research use that costs about $50 in prototype form.

Advances in neurological monitoring

New technologies for in vivo neurological monitoring also were described at the conference. There is a small but well-established market for products used in neurological monitoring, including devices for tracking intra-cranial pressure, EEG equipment, and cerebral oxygenation monitoring products. As shown in Table 3, the U.S. market for products in those three segments totaled about $165 million in 2002, growing at around 10% annually.

One of the most promising new monitoring technologies, optical topography, was described by Richard Kennan, PhD, of Albert Einstein College of Medicine (New York). Optical topography was developed by Hitachi Medical (Tokyo) and was recently recognized as one of the most important inventions of 2002. The Hitachi system, the ETG-100, uses near-infrared spectroscopy and a high-density array of 1 mm diameter optical fibers to non-invasively map and monitor concentrations of oxy- and deoxy-hemoglobin, cytochrome oxidase and myoglobin on the cortical surface. Others have used near-infrared spectroscopy to monitor global brain oxygenation changes associated with hypoxia and neural activity.

The Hitachi system operates at 780 nm and 830 nm, shining light through optical fibers that are positioned on the surface of the patient's head. Other fibers spaced 3 cm from the emitting fibers are used to detect the reflected and scattered light. The measurement does not require shaving the head, although there is a limitation on patient mobility. Spatial resolution is now about 25 mm, although Kennan said he believes it may be possible to reach 10 mm resolution with further development of the technology, perhaps employing time-of-flight measurement methods.

Because of the lack of electrical signals, the optical topography system can easily be integrated with EEG monitoring without concerns about electrical interference, allowing molecular distributions in the cerebral cortex to be mapped while simultaneously monitoring EEG signals over the same region of the brain.

The optical topography system has been used for presurgical evaluations, e.g., for surgical planning, by allowing neurologists to monitor brain response during various types of mental activity at relatively low cost as compared to techniques such as functional MRI. In fact, studies conducted by Kennan have shown a good correlation between optical topography measurements and functional MRI images, but there is no need for injection of tracers as with functional MRI.

Other applications include monitoring of neonates to assess brain activity during learning (e.g., as speech capabilities develop) and stroke rehabilitation. Current development efforts are aimed at systems for whole-brain monitoring, and at improving spatial resolution. Future applications are expected to expand to include monitoring of patients with multiple sclerosis, Parkinson's disease, Alzheimer's disease, dyslexia, attention deficit disorder and sleep apnea.

Invasive monitoring technologies are also of interest for neurological applications, although at least for now such methods are reserved mainly for animal research. Mark Wightman of the department of chemistry and neuroscience center at the University of North Carolina (Chapel Hill, North Carolina) described an implantable microelectrode technology using cyclic voltammetry to monitor changes in dopamine concentration in the brains of rats during various types of physical and mental activity. While the use of invasive electrodes is unlikely to be implemented on a wide scale in routine clinical practice solely for monitoring, the techniques could potentially be used along with a variety of implantable electrical stimulation devices to monitor brain activity and perhaps provide feedback control of such devices. The primary application, however, is in research studies of brain response to various stimuli. Studies have also been performed using the technology in animal models of Parkinson's disease.