Outcomes Assessment in Future Clinical Trials
This report was reviewed for medical and scientific accuracy by Andrew Goodman, MD , Department of Neurology, University of Rochester Medical Center, New York.
Future clinical trials of agents targeted for multiple sclerosis (MS) will increasingly rely on objective outcome measurements that image central nervous system lesions, according to research scientists who discussed ways to improve clinical trials at the International Business Communications 6 Symposium on Multiple Sclerosis.
Particularly exciting is the prospect of expanded use of magnetic resonance imaging (MRI), including novel MRI techniques, to track the evolution of lesions. This approach should be an effective way of observing the process of demyelination and lesion formation over time and, ideally, the benefits of treatment on remyelination. MRI techniques will also allow pilot studies to be completed more quickly, with fewer subjects, experts predicted.
Fred D. Lublin, MD, Professor and Vice-Chairman Department of Neurology, Neurology Service Chief, Medical College of Pennsylvania Hospital and Director of the Multiple Sclerosis Center, MCP Hahnemann University, said MRI is gaining value in establishing disease activity and lesion load and will play an important role in future trials as well as clinical practice. "MRI is a very sensitive tool. While MRI findings have been used as back-up information and secondary outcome measures in trials, the MRI data has often been the swaying factor in the FDA approval process," he pointed out.
"I believe that MRI shows the pathology of MS. Not all lesions seen on MRI show up clinically, but ultimately these white spots will affect patients. While cerebral MRI abnormalities do not correlate well with disability scores, they do correlate with neuropsychological tests," he said.
Standard clinical parameters-relapse rates and disease progression assessment- are the currently employed measures of disease activity, but they are less-than- ideal methods of gauging changes in the patient's disease state, although these parameters are used as primary outcome measures in clinical trials. For example, while all three currently available agents (interferon beta-1b, interferon beta- 1a, and glatiramer acetate) reduce relapse rate by about one third. Patients in placebo arms perform well in clinical trials, which complicates the determination of treatment efficacy, said Dr. Lublin, Chair of the National MS Society's Advisory Committee on Clinical Trials.
There are additional problems with these outcome measures. The relapse rate in relapsing remitting MS tends to diminish over time, and regression to the mean occurs-that is, patients selected because of their frequent exacerbations gradually come to resemble the typical patient.
Furthermore, the Kurtzke Expanded Disability Status Scale (EDSS) has inherent flaws, Dr. Lublin added. This ordinal, non-linear scale ranges from "trivial to severe," the half-point intervals do not carry equal weight across the scale, and changes that are documented can be biased when evaluations are performed during exacerbations, Dr. Lublin suggested.
The recent PRISMS (Prevention of Relapses and Disability by In terferon - beta- 1a Subcutaneously in Multiple Sclerosis) study used a disability instrument called the integrated disability status score (IDSS). The IDSS measures the total degree of disability from exacerbations and progression of disease, yielding the total treatment effect on the disease. But a better instrument yet may be the Multiple Sclerosis Functional Composite (MSFC),a new scale developed by a task force of the National MS Society. The Multiple Sclerosis Composite Functional Score is a non-physician based continuous outcome measure that demonstrates changes in disability by determining standard deviations above and below the mean. It encompasses the timed 25-foot walk, the 9-hole peg test, and the paced auditory serial addition test, averaging the scores to produce a final assessment (Z score).
"We are hopeful that this will be a better outcome measure because it is faster and requires fewer patients to see results," Dr. Lublin offered.
The eventual acceptance of more objective endpoints will follow other advances in clinical trial design, he pointed out. Most notably, the definitions of the MS disease types were standardized only a few years ago, and several terms, including chronic progressive and relapsing progressing MS, were eliminated. "This is an exciting decade for MS clinical research. There have been several successful clinical trials and a sufficient number of negative ones to keep us honest," Dr. Lublin commented. "We have improved upon trial design with each study, and we now have 'partial therapies,' with room for improvement. What we are trying to do now is improve our ability to decrease inflammation and demyelination, decrease axonal loss, and increase repair. These will be major frontiers."
New Imaging Assessments
Nancy Richert, MD, PhD, IPA Fellow, Laboratory of Diagnostic Radiology Research at the National Institutes of Health and Assistant Clinical Professor of Radiology at Children's National Medical Center and George Washington University Medical Center, described MRI's ability to show that substantial disease activity occurs in the setting of clinically quiescent disease. Serial monthly gadolinium enhancements show four to ten times more disease activity than can be demonstrated by clinical scores and this is even considered to be an underestimate of lesion volume, she reported.
"If you triple the dose of contrast, you get a 70% increase in lesion frequency. If you delay the scan for 20 to 40 minutes after injection, you see more lesion development. And if you use a new technique called magnetization transfer (described below), you get even more. With the combination of these methods, you get a 130% increase in the number of lesions seen on MRI," Dr. Richert said. "So typically we are not seeing all the lesions that are there. But there is not enough natural history data to know if the undetected lesions correlate with disease."
Furthermore, lesion enhancement is highly variable, both between patients and, in a given patient, from month-to-month. In a study of 135 relapsing remitting patients, the mean number of new enhanced lesions was 3.3 per month, but some patients had as many as 50-75, she reported.
MRI offers the opportunity to observe this evolution of lesions, which are gaining importance as disease markers. With g adolinium enhancement, one can image the initial event--the breaking down of the blood brain barrier--which persists for about one month, then leaves a T2 hyperintense lesion. This T2 lesion is quite heterogeneous, representing a whole spectrum of inflammation that can include demyelination, axonal loss, and remyelination. One of the shortcomings of T2 imagery, however, is its inability to demonstrate what stage the lesion is in or how long it has been there. But T2 lesions can be quantitated to produce the total lesion load or volume, although this, too, is variable because of the cycle of inflammation and resolution. In clinical trials, this information is used to judge a drug's effect on the accumulation of T2 lesions over time, according to Dr.Richert.
Dr. Richert added that MRI abnormalities may be most effectively appreciated during relapses, when the total number and volume of lesions increases dramatically, both on T2 images and gadolinium enhancement. These findings unfortunately, correlate poorly with EDSS scores. The correlation improves, however, if one looks only at the "black holes" (areas of severe tissue destruction), and they correlate with neuropsychological tests.
While MRI directly visualizes the brain, its lack of correlation with EDSS scores is a primary reason why abnormalities on MRI are not considered the perfect surrogate markers, and why they are not accepted as primary outcome measures in clinical trials, she said. But MRI has advantages over the EDSS because it is more objective, measures total CNS disease, demonstrates subclinical disease, is a fixed document, entails less interobserver variability, can be statistically evaluated, and requires fewer patients and a shorter time period to show a treatment effect, Dr. Richert pointed out.
Because MRI criteria are very sensitive, MRI assessments are desirable for clinical trials. But MRI data are used as primary outcome measures only in phase I or II trials and as secondary outcome measures in phase III trials, where the information is obtained as changes in T2 lesion load at 6 to 12-month intervals. The lesion accumulation can be measured at these intervals to get an idea of a drug's efficacy, Dr. Richert said.
In fact, NIH trials have shown rapid treatment effects of bet a-interferon in decreasing gadolinium-enhanced lesions and, correspondingly, T2 lesion loads. The total number of new or enhanced lesions diminishes more dramatically with gadolinium enhancement than with T2 lesion load, because T2 volume includes a permanent residual abnormality, she added.
A number of new MRI measures are increasing the sensitivity and specificity of this MRI, including magnetization transfer imaging, spectroscopy, and FLAIR (fluid attenuated inversion recovery) technology. These assessments should become very important in future clinical trials Dr. Richert predicted.
Magnetization transfer is being employed to visualize demyelination and remyelination. This technique measures the exchange between water that is bound to myelin and the free water in tissue, producing a ratio that conveys whether the lesion represents an acute inflammatory process or the healing process. The ratio remains low, for example, when the lesion is not recovering, and the longer the disease and the greater the lesion loads, the lower the magnetization transfer ratio, Dr. Richert explained.
FLAIR may enhance the sensitivity of T2-weighted measurements by nulling the cerebrospinal fluid so that the background CSF appears dark rather than bright, making for an easier and more accurate quantitation of lesions. To image axonal loss, spectroscopy is being used, demonstrating a decrease in NAA in lesions (probably a neuronal marker) and an increase in choline (representing membrane breakdown); magnetization transfer correlates well with spectroscopy and may, therefore be an indirect measure of axonal loss, she said.
In an interview Dr. Richert added that these MRI techniques can not only add important information to research studies, but can be enormously helpful to physicians caring for MS patients now. To make MRI more reproducible, and therefore more useful in serial assessments, she recommended that technicians place oil capsules, which appear bright on imaging, in patients' ears to aid in localization and positioning techniques.
The Multiple Sclerosis Express Report(tm): Outcomes Assessments in Future Clinical Trials discusses data that was presented at a Pre-Conference Workshop at the IBC's 6th International Symposium on Multiple Sclerosis on December 10, 1998. Other presentations of interest include:
1 "Th1 and Th2 Cytokine Analysis in MS Patients Using the ELISPOT Assay," Pelfrey, C,Ph.D., Cleveland Clinic Foundation.
2 "Quinolinic Acid in MS: Surrogate Marker and Endogenous Toxia," Reinhard Jr., J, Ph.D., Senior Research Investigator, MolecularPharmacology, Glaxo Wellcome.
3 "Characterization o f MS and normal brain diff erences using global magnetization transfer," Ostuni JL, Richert ND, Lewis BK, and Frank JA. In press.AJNR. American Journal of Neuroradiology (1998).