Abstract
Objective: To determine if progressive brain atrophy could be detected over 1- and 2-year intervals in relapsing MS, based on annual MR studies from the Multiple Sclerosis Collaborative Research Group (MSCRG) trial of interferon β-1a (Avonex).
Methods: All subjects had mild to moderate disability, with baseline expanded disability status scores ranging from 1.0 to 3.5, and at least two relapses in the 3 years before study entry. Atrophy measures included third and lateral ventricle width, brain width, and corpus callosum area.
Results: Significant increases were detected in third ventricle width at year 2 and lateral ventricle width at 1 and 2 years. Significant decreases in corpus callosum area and brain width were also observed at 1 and 2 years. Multiple regression analyses suggested that the number of gadolinium-enhancing lesions at baseline was the single significant contributor to change in third ventricle width. Atrophy over 1 and 2 years as indicated by enlargement of the third and lateral ventricle and shrinkage of the corpus callosum was greater for patients entering the trial with enhancing lesions. Greater disability increments over 1 and 2 years were associated with more severe third ventricle enlargement.
Conclusion: In patients with relapsing MS and only mild to moderate disability, significant cerebral atrophy is already developing that can be measured over periods of only 1 to 2 years. The course of cerebral atrophy in MS appears to be influenced by prior inflammatory disease activity as indicated by the presence of enhancing lesions. Brain atrophy measures are important markers of MS disease progression because they likely reflect destructive and irreversible pathologic processes.
Several types of measures based on MRI are now used to evaluate and quantify the impact of MS on the CNS, including the number or volume of gadolinium-enhancing lesions and the T2-hyperintense (T2) lesion load. The former serves as an objective assay of acute and active disease, whereas the T2 lesion load, or number of new or enlarged T2 lesions, provides a measure of past history, because each acute event leaves a nonspecific T2 “footprint.” These MR-based measures are routinely used as outcome measures in phase III treatment trials in Europe and North America.1,2
Despite widespread acceptance of MR measures as generally reflecting ongoing and past brain insult, no measure or combination of measures is recognized as providing a complete description of the types of damage that are the natural consequence of MS. T2 lesions are believed to lack specificity because they reflect a number of pathologic processes, ranging from edema and mild to moderate demyelination through astrogliosis and axonal injury or loss.3-5 The correlations between the MR-measured lesion burden and clinical status are usually poor. For example, T2 lesion loads have been shown in large trials to correlate only poorly with disability.6-8
Motivated in part by the need for more efficient, sensitive, and clinically relevant surrogate end points in assessing treatment efficacy in MS, several additional MR-based measures of CNS damage are under evaluation. These include the magnetization transfer index,9,10 T1-hypointense lesion volume,11,12 concentration of the neuronal marker N-acetylaspartate (NAA) as seen by proton MRS,13 and measures of cerebral or spinal cord atrophy.14-17 The clinical significance of CNS atrophy has only recently come under investigation in longitudinal trials, and the role of atrophy measures in formal trials remains undetermined.
To learn more about the natural history of brain atrophy in MS, a prospective study was conducted in patients with relapsing MS who participated in a phase III trial evaluating interferon β-1a (Avonex; Biogen, Cambridge, MA).18,19 The goal of this study was to determine if progressive brain atrophy could be detected at 1- and 2-year intervals in patients with relapsing MS, and to assess the relationship between atrophy measures and standard MR and clinical outcome variables.
Methods.
Patients.
The results reported here are for untreated patients (pre-randomization and placebo-treated) from the phase III Multiple Sclerosis Collaborative Research Group (MSCRG) trial. Studies of atrophy as a potential treatment outcome measures will be the subject of a separate publication. Trial design, eligibility and entrance criteria, and primary outcomes have been reported.8,18,19 A total of 301 patients with relapsing-remitting MS were enrolled at four clinical sites (Buffalo, NY; Cleveland OH; Washington, DC; Portland, OR) and randomized to receive either placebo or 30 μg (6 million IU) of interferon β-1a (Avonex) once weekly by intramuscular injection.18,19 The number of subjects with follow-up MR studies was less than the number initially enrolled primarily as a result of the decision, based on a lower than originally estimated dropout rate, to end the study early.19 Patient demographics are summarized in table 1. Before study entry, all subjects were required to have had at least two exacerbations over the previous 3 years, and not to have experienced an exacerbation within the 2 months before study entry. All patients had baseline Expanded Disability Status Scale score (EDSS) of 1.0 to 3.5, inclusive. The primary outcome variable was the time to onset of sustained worsening in disability; secondary outcome variables included exacerbations and changes on MRI.
Baseline clinical and MR characteristics for 3 patient groups
MR imaging.
The MRI protocol was based on high-field imaging studies conducted at baseline and at 1- and 2-year intervals. MRI was performed each year according to a standardized protocol.8,18,19 The protocol included 1) a 5-mm–thick sagittal T1-weighted series (TR/TE 600/20) with a 192 × 256 matrix, a 24-cm field of view, and one excitation; 2) an axial 5-mm interleaved (nongapped) dual-echo, spin-echo series acquired with intermediate and T2-weighting (TR 2,000, TE 30, 90 msec) through the brain, with a 192 × 256 matrix, a 24-cm field of view, and one excitation; and 3) a precontrast- and postcontrast-enhanced, T1-weighted, axial spin-echo series with TR/TE 600/20, with otherwise similar parameters. Contrast (gadolinium-chelate) was injected intravenously at a dose of 0.1 mmol/kg over a 1- to 2-minute period to a maximum volume of 20 mL, with the postcontrast T1-weighted series starting 5 minutes after injection.
MR variables.
T2 lesion volume, atrophy measures, and gadolinium-enhancing lesion number and volume were obtained at baseline (before administration of interferon β-1a or placebo) and at 1 and 2 years thereafter. The numbers of new and enlarging T2 lesions were determined at year 2 with comparisons against baseline scans.8 T2 lesion volumes were based on manual tracings on a computer workstation, with all analyses for an individual patient performed on a single day as a cluster.8 Gadolinium-enhancing lesion counts and volumes have been reported previously for this study population.19 MR results were determined at the University of Colorado Health Sciences Center (Denver, CO), where analyzers were blinded to patients’ treatment status.
Atrophy measures.
Atrophy measures, in duplicate, were executed by one neuroradiologist (J.H.S.) over a 4-year period and were based on the precontrast axial (ventricle and brain width) and sagittal (corpus callosum) T1-weighted series (figure 1). Corpus callosum area was determined by outlining the margins of the structure on a high-resolution workstation monitor.17 Third ventricle and lateral ventricle widths were determined along a plane corresponding to the (anteroposterior) midpoint of the ventricle. Brain width was considered to be the distance between two points on the cortical surface, measured at the same level as lateral ventricle width. Intraobserver measurement error was determined by calculating the coefficient of variation as a percentage (SD/mean) for two measurements made at a test–retest interval of 8 to 13 days by a single observer (J.H.S.) in 10 random cases. The mean coefficients of variation were 7%, 4%, 1%, and 3% for third ventricle width, lateral ventricle width, brain width, and corpus callosum area, respectively. The coefficient of variation for third ventricle decreased to 3%, excluding the one test case with mean width ≤1.5 mm.
Figure 1. Atrophy measures based on T1-weighted images. (A) Arrows indicate level and location used for third ventricle width measure. (B) Maximum lateral ventricle width (arrows) determined at a level at which septum pellucidum remains thin. (C) Brain width (arrows) determined at same level as in panel B. (D) Corpus callosum (arrows) traced using best available mid-sagittal section.
Statistical methods.
Within-patient change from baseline atrophy measures was tested using a one-sample t-test. The Wilcoxon rank sum test was used for between-group comparisons. Correlations were determined based on the Spearman rank statistic. All analyses were performed as two-tailed tests. A forward stepwise regression model was used to identify patient characteristics or characteristics of the natural disease course that were predictive of change in the atrophy measures. Variables that were used as potential predictors of time to disability progression were used in each of the models, as described previously.20
Results.
Correlations at baseline.
Correlations between measures of brain atrophy and other study variables at baseline are presented in tables 2 through 4⇓⇓. As shown in table 2, there were modest correlations between most pairs of atrophy measures, the strongest between third ventricle and lateral ventricle width (r = 0.53, p = 0.0001). The third and lateral ventricle width was negatively correlated with corpus callosum area (r = −0.40 and −0.25, respectively; p = 0.0001).
Correlations between brain atrophy measures at baseline
Correlations between brain atrophy measures and T2 lesion volume and number and volume of gadolinium-enhancing lesions at baseline
Correlations between baseline MR measures and age, Expanded Disability Status Scale (EDSS) score, duration of MS, and pretrial exacerbation rate
Table 3 summarizes correlations at baseline between the brain atrophy measures and several standard MR-based measures of disease. Moderate correlations (‖r‖ = 0.39 to 0.49, p = 0.0001) were seen between T2 lesion volume and three of the atrophy measures (third and lateral ventricle widths, and corpus callosum area). There was either no correlation or at most a weak correlation (for lateral ventricle width) between the atrophy measures and baseline enhancing lesion number or volume.
Table 4 summarizes correlations between MRI measures and clinical characteristics at or before the baseline study. Small but significant correlations with baseline EDSS were seen for the third (r = 0.26, p = 0.0001) and lateral ventricle width (r = 0.18, p = 0.0025), comparable with the correlation between baseline EDSS and T2 lesion volume (r = 0.22, p = 0.0007). The correlation between EDSS and corpus callosum area was also small, but significant (r = −0.15, p = 0.016). The correlations between duration of MS and third ventricle width, lateral ventricle width, corpus callosum area, and T2 lesion volume were all small but significant.
Longitudinal analyses.
The results for within-subject changes in brain atrophy measurements at 1- and 2-year intervals are shown in figure 2. Increases in lateral ventricle width were significant at both the 1- and 2-year measurement point (p = 0.0001 for both years), and in third ventricle width at 2 years (p = 0.027). A significant decrease was seen for corpus callosum area at the year 1 and 2 measurement points (p = 0.0001 for both years), and for brain width at 1- (p = 0.016) and 2-year intervals (p = 0.0001). For third ventricle and lateral ventricle width, the mean increase was 4.5% and 5.5% per year, respectively (figure 3). The corpus callosum area and brain width decreased an average of 4.9% and 0.64% per year, respectively (see figure 3).
Figure 2. Mean within-subject changes in untreated patients at 1 and 2 years. Changes at 1 year based on 124 paired sets of MRI studies; 2-year data from 85 paired sets of MRI data. *p < 0.05; **p = 0.0001. Bars represent standard error of the mean. 3V = third ventricle width; LV = lateral ventricle width; CC = corpus callosum area; BW = brain width.
Figure 3. Mean values for untreated patients for four atrophy measures (A, third ventricle; B, lateral ventricle; C, corpus callosum; D, brain width) at baseline (year 0), year 1, and year 2. Analyses based on 85 patients followed by MRI for 2 years. Error bars represent the standard error of the mean.
Factors contributing to brain atrophy.
Multiple regression analyses revealed that the number of enhancing lesions at baseline was the single significant contributor (p ≤ 0.001) to brain atrophy of all factors entered into the model. The higher the number of enhancing lesions at baseline, the greater the atrophy over 2 years. The partial model r2 was 0.192 for changes in third ventricle width. Other variables, including EDSS, disease duration, and T2 lesion volume, did not appear to contribute significantly to the model.
Table 5 shows within-subject changes at years 1 and 2 for the four atrophy variables in the patients with one or more enhancing lesions at baseline (GD+), compared to those without enhancing lesions (GD−). Within-subject increases in third ventricle and lateral ventricle width were greater at years 1 and 2 for patients with one or more enhancing lesions at baseline (GD+) compared with those without enhancing lesions at baseline (GD−). Similarly, within-subject decreases in corpus callosum area were greater for the GD+ group. Differences between the GD+ and GD− groups were significant for third ventricle at years 1 and 2 (p = 0.026 and p = 0.027, respectively), and for corpus callosum at year 1 (p = 0.046). Similar effects of enhancing lesion status were seen for the GD+ versus GD− groups for change in T2 lesion volume.8
Mean (SD) within-subject change in atrophy variables for patients with (GD+) and without (GD−) gadolinium-enhancing lesions at baseline
Table 6 summarizes correlations between the change in third ventricle width and the change in other MRI and clinical parameters. For third ventricle width, there were small (r = 0.21 to 0.38) but consistent and significant correlations for changes in width and enhancing lesions, new and enlarging T2 lesions, and T2 lesion volume increment.
Summary of on-trial correlations between change from baseline in third ventricle width and other study variables
There was only a weak trend for the correlation between increase in third ventricle width and EDSS at years 1 (r = 0.16, p = 0.075) and 2 (r = 0.20, p = 0.065), which was slightly stronger in the group of patients with enhancing lesions at baseline (r = 0.22, p = 0.07 and r = 0.35, p = 0.018, respectively). Figure 4A shows the relationship between disability progression and atrophy based on third ventricle width for patients with greater versus lesser EDSS increments. Patients with greater increments in disability have greater atrophy around the third ventricle compared with the less disabled cohort. Similar trends (not significant) were seen for lateral ventricle and disability, whereas the changes in corpus callosum area and brain width showed no relationship to disability. For this same population, generally weak trends or no differences were seen for change in T2 lesion volume for the multiple EDSS increment groupings (see figure 4B).
Figure 4. Analyses of third ventricle atrophy and T2 lesion volume change in untreated patients with multiple Expanded Disability Status Scale score (EDSS) progression cutoff points. The figure shows the mean 1- and 2-year increases in third ventricle width (A) and median changes in T2 lesion volume (B) for the subgroups showing either greater or lesser EDSS increments. EDSS values are based on sustained increases (over a minimum of 6 months), as described previously.18,19 p Values for comparisons above (gray bars) versus below (black bars) cutoff points based on t-test (third ventricle) and Wilcoxon rank sum test (T2 lesions). Numbers below bars are the sample sizes for each group. For third ventricle, error bars are the standard error of the mean.
Discussion.
The results from this study show that patients with relapsing MS and only mild to moderate disability (EDSS 1.0 to 3.5) are already in the process of development of significant cerebral volume loss that is quantifiable over a 1- and 2-year observation period. Furthermore, the data suggest a relationship between atrophy and disease activity as measured by enhancing lesion activity.
Prior cross-sectional studies have shown that patients with MS are more likely than matched normal control subjects to have tissue loss, including in the corpus callosum17,21-24 and the spinal cord,16,25-28 and based on total cerebral volume measures.29 Atrophy also occurs in the cerebellum in MS,14 and the lateral and third ventricles can enlarge, with the latter frequently apparent by visual inspection alone. Ventricular enlargement is thought to be the result of volume loss in the adjacent tissues (ex vacuo enlargement). Longitudinal studies of CNS atrophy have been limited. In one recent study,15 27 patients with moderate to severe relapsing-remitting or secondary progressive disease (EDSS 3 to 7) were followed for 18 months. Sixteen of these patients showed a decrease in the area of the four central brain slices that were evaluated. In another study of 29 patients with relapsing and progressive MS, followed for a mean period of 12 months, a significant decrease in cervical spinal cord area was detected.28
Our longitudinal study showed that for all four prospectively selected atrophy measures used in the MSCRG trial, changes were apparent at the 1- and 2-year observation intervals. These changes occurred in the predicted direction, consisting of an increase in the width of the third and lateral ventricles and a decrease in corpus callosum area and brain width. The magnitude of the changes based on third and lateral ventricles and corpus callosum was on the order of 5% per year. These impressive central brain atrophy rates are based on observations over a relatively narrow time window (2 years) in a specifically defined population of patients with relapsing MS. Further studies would be required to determine if these rates are representative of other stages of disease.
The atrophy measures show modest correlations between themselves (e.g., third ventricle width with lateral ventricle width and corpus callosum area), suggesting that the processes leading to atrophy may have both a global and local anatomic impact. Three of the atrophy measures (third ventricle and lateral ventricle width, corpus callosum area) also correlate reasonably well with the MR measure of cumulative past insult, the T2 lesion volume at study entry (‖r‖ = 0.39 to 0.49), and less strongly, but significantly, with disease duration before study entry.
Most cross-sectional studies have suggested that there may be a relationship between atrophy and neuropsychological dysfunction. The width of the third ventricle in one CT study appeared to be the best indicator of intellectual and memory dysfunction.30 By MRI, atrophy of the corpus callosum has been associated with interhemispheric dysfunction23,24,31 and dementia.32 The correlation between atrophy measures and neuropsychological dysfunction has not yet been evaluated for patients in this study, but will be the subject of future publications. Several additional studies have found relationships between atrophy and impaired function and disability. Volume loss in the cerebellum is associated with poor cerebellar function,14 and spinal cord atrophy and disability based on the EDSS appear to be well correlated.16,27,28 In a longitudinal study, atrophy of central brain slices was significantly greater in the subset of patients with a sustained deterioration in EDSS.15 In this study, three of the four atrophy measures (third and lateral ventricle width, corpus callosum area) correlate significantly with EDSS at study entry. For the third ventricle, where the strongest and most consistent relationships are seen, the correlation with EDSS (r = 0.26, p = 0.0001) was similar to that observed for T2 lesion volume and EDSS at baseline (r = 0.22, p = 0.0007) in this study8 and another large, prospective trial.7 Our on-trial results, however, point to an interesting relationship between third ventricle atrophy and progressive disability, which is stronger than the relationship between T2 lesions and disability.
There were modest correlations (r = 0.21 to 0.38, p = 0.0004 to 0.03) for change in third ventricle width with standard MR measures that were used for disease activity, including increase in T2 lesion volume, new and enlarging T2 lesions, and cumulative enhancing lesion number or volume. This is not entirely surprising because a case can be made for a simple linkage in the natural history of new, individual lesions that ultimately leads to cerebral atrophy.2,33 The enhancing lesions, the earliest detectable MR evidence for new lesion formation, correlate with the inflammatory stage of lesions seen on histopathologic study.4 Later stages in lesion evolution culminate in a chronic T2 lesion, which is basically a permanent marker or footprint of the prior inflammatory and demyelinating insults to the brain. The T2 lesion has been described as nonspecific because it is characterized by variable levels of tissue damage, ranging from edema through demyelination, astrogliosis, and axonal injury and loss. Demyelination and axonal loss collectively would be expected to result in both a local volume loss and a more diffuse cerebral atrophy.
The presence of enhancing lesions at baseline predicts a greater rate of atrophy. This putative relationship is best measured by increase in third ventricle width, but similar trends are seen for lateral ventricle and corpus callosum–based measures. This relationship between enhancing lesions (presumably reflecting inflammatory activity) and cerebral atrophy is seen even with only a single baseline observation (an MR “snapshot”), but might become even stronger if multiple baseline MR observations and more sensitive measures of inflammation were available. Enhancing lesion activity also predicts a greater accumulation of T2 lesions over time (measured as either new or enlarging lesions), an increase in T2 lesion volume, and subsequent gadolinium-enhancing lesions and clinical relapses.8,33-35 Although we believe that the relationship between enhancing lesions and atrophy is likely causal, other factors cannot be excluded as potentially contributory. Most important, endogenous and exogenous corticosteroids have been associated with both focal (typically hippocampus) and global, reversible and irreversible brain volume loss.36,37 The trial protocol permitted only short-term administration of corticosteroids for disease exacerbations.18 Patients with active disease as indicated by enhancing lesions or clinically evident exacerbation are precisely those most likely to receive corticosteroids. Consequently, our study design does not exclude the possibility that corticosteroids could contribute to brain atrophy. However, in MS, corticosteroids are administered to decrease the severity of clinical symptoms thorough anti-inflammatory and immunomodulatory mechanisms.38 It is therefore reasonable to hypothesize that the effect of such therapy would be, if anything, to limit subsequent tissue damage.
Because there are few correlative imaging–neuropathology studies, little is known about the relative contributions of the various factors responsible for cerebral volume loss underlying these atrophy measures. Volume loss in MS is likely the result of myelin and axonal loss, and there may be contributions from reductions in axonal diameter39 and tissue contraction from astrogliosis. In vivo proton MRS studies provide evidence for axonal or neuronal damage in MS based on decreased levels of NAA.13,14 The factors underlying atrophy may vary anatomically. For example, lateral ventriculomegaly may result from adjacent demyelination because T2 lesions are often extensive surrounding the lateral ventricles. However, MR-detectable white or gray matter lesions are relatively sparse around the third ventricle. It is therefore possible that volume loss indicated by an increase in third ventricle width may be the result of distant insults, with secondary volume loss from axonal and wallerian degeneration. Alternatively, MR may be insensitive to gray matter or other small but no less functionally important microscopic demyelinated lesions in tissues surrounding the third ventricle. Although there is no direct evidence, the third ventricle is well positioned to reflect damage in distant anatomic regions because numerous afferent and efferent projections course through the thalamus.40 Recent studies suggest a strong correlation between axonal injury and acute, inflammatory lesions in MS.39 As this study has shown, enhancing lesions predict several future outcomes, including atrophy. The pathologic changes that lead to atrophy presumably are initiated early in the course of disease at the time of acute, enhancing lesion activity.
Patients with greater disability progression show larger, significant increments in atrophy as indicated by third ventricle measurements. In a prior series,15 a significant change in EDSS was seen in the subgroup of patients defined as showing progressive atrophy based on brain slice area measures. The relationship between structural damage and disability as indicated by the EDSS is not likely to be straightforward. Small lesions in functionally sensitive locations are likely to be disproportionately influential on the EDSS or other rating scales, yet would not be expected to result in the more diffuse injuries presumed to be responsible for detectable volume changes and atrophy. How the atrophy measures compare with other, more tissue damage–specific measures, such as those based on magnetization transfer,9,10 T1 relaxation,11,12 or the neuronal marker NAA,13 is unknown. Several pilot and preliminary studies suggest that the “more specific measures” may be more strongly correlated with disability compared with T2 lesion volume. Magnetization transfer, T1 lesions, and NAA measures have not as yet been formally evaluated by large, prospective, longitudinal studies or trials.
For this study, which was designed in the late 1980s, we used simple computerized but manual linear or area measurements, based on MRI with 5-mm interleaved (no gap) slice profiles. The resultant data are almost certainly strongly affected by slice location and partial volume effects. Future efforts can be directed toward improving the methods of quantifying both focal and global atrophy–for example, using acquisitions with submillimeter dimension three-dimensional MR data sets, and automated computer-based measurement methods. Despite several limitations, our measures, particularly those based on third ventricle width, show that important dynamic changes are already set in motion in the brain of these patients with relapsing MS and only mild to moderate disability.
The association between gadolinium-enhancing lesions and more active brain deterioration, indicated, for example, by enlargement of the third ventricle and shrinkage of the corpus callosum, suggests that treatments that decrease enhancing, presumably inflammatory MS lesions, such as the β-interferons,19,41,42 also should slow the cumulative pathologic process that results in atrophy.
Although atrophy was used as an exploratory outcome measure in this treatment trial, the results suggest that further investigation of atrophy-based measures is warranted as disease classification, prognostic indicators, and potential treatment trial outcome measures. Future studies should be directed to defining further the natural history of these changes, their relationships to disability and neuropsychological impairment, and determining the impact of treatment.
Appendix
The Multiple Sclerosis Collaborative Research Group (MSCRG) consists of the following sites and their respective study personnel, in addition to the cited authors:
Buffalo, NY—William C. Baird Multiple Sclerosis Research Center, Millard Fillmore Health System: P.M. Pullicino, MD, PhD, L.M. Bona, M.E. Colon-Ruiz, BS, N.A. Donovan, RN, S.B. Illig, RN, MS, NP, Y.M. Kieffer, RN, BSN, M.A. Umhauer, RN, MS, NP; Department of Neurology, The Buffalo General Hospital: C.E. Miller, RN, NP, DNS; Division of Developmental and Behavioral Neurosciences, Department of Neurology, The Buffalo General Hospital: A.K. Kilic, MS, E.L. Sargent, BS, M. Schachter, PhD, D.W. Shucard, PhD, V. Weider, PhD; Physicians Imaging Center of Western New York: B.A. Catalano, RT, J.M. Cervi, RT, C. Czekay, RT, J.L. Farrell, RT, J.S. Filippini, RT, R.C. Matyas, RT, K.E. Michienzi, RT; Department of Microbiology, Roswell Park Cancer Institute: M. Ito, MD, J.A. O’Malley, PhD; Department of Social and Preventive Medicine, School of Medicine and Biomedical Sciences, State University of New York at Buffalo: C.V. Granger, MD, M.A. Zielezny, PhD; MSCRG Data Management and Statistical Center, Department of Neurology, The Buffalo General Hospital: R.L. Priore, ScD, J.M. Brun, BS, L.A. Green, RRA, BS, K.M. O’Reilly, MS, J.A. Shelton, MS.
Cambridge, MA—Biogen, Inc.: John J. Alam, MD.
Cleveland, OH—Mellen Center for Multiple Sclerosis Treatment and Research, Cleveland Clinic Foundation: B. Weinstock-Guttman, MD, D.Y. Barilla, MA, S.L. Boyle, BS, K.K. Perkins, BA, J.E. Perryman, B.G. Stiebeling, RN, MSN; Department of Diagnostic Radiology, Cleveland Clinic Foundation: J.F. Konecsni, RT, J.S. Ross, MD.
Denver, CO—Department of Radiology–MRI, University of Colorado Health Sciences Center: K. Choi, PhD, C.J. Gustafson, RT, D. Singel RT, D. Lindsey RT, K. McCabe BS, M. Meyer, MS, B.J. Quandt, R. Leek, BS, A.L. Scherzinger, PhD, J. Sheeder, BS.
Portland, OR—Department of Neurology, Good Samaritan Hospital and Medical Center: M.K. Mass, MD, D.A. Griffith, RN; Department of Neurology, Oregon Health Sciences University: D.N. Bourdette, MD, R.H. Whitham, MD, J.M. Harris, BS, M.D. Lezak, PhD, I. Mimica, PhD, J.A. Saunders, RN, ANP; Department of Radiology, Good Samaritan Hospital and Medical Center: W.E. Coit, MD, C.R. Force, RTR, F.J. Gilmore, RTR, L.B. Harris, RTR, M.M. Jones, MD, J.A. Kauffman, RTR, K.E. Marberger, RTR, J.W. McBride, RTR, L.L. Miller, RTR, G.K. Wright, RTR.
Washington, DC—Department of Neurology, Walter Reed Army Medical Center: D.M. Bartoszak, MD, J. Braiman, MD, M.E. Coats, MD, D.S. Dougherty, MD, J.A. Brooks, RN, MSN, H.R. Brown, M.E. Graves, RN, J.A. Schmidt, RN, DNSc; Department of Neurology, Georgetown University Medical Center: S.L. Cohan, MD, J.W. Mothena, BSN, RN; Cognitive Neuroscience Unit, National Institute of Neurological Disorders and Stroke, NIH (Bethesda, MD): J.H. Grafman, PhD, M.K. Kenworthy, BA, M.M. Morton, BS, MEd; Department of Radiology, Walter Reed Army Medical Center: D.M. Brown, RT, D.C. Brown, MD; Department of Radiology, Georgetown University Medical Center: L.M. Levy, MD, PhD.
Springfield, VA—Department of Neurology, Kaiser Permanente Medical Center: B.J. Scherokman, MD.
The following scientific consultants were involved in the planning of this study: University of Maryland Cancer Center (Baltimore, MD): E.C. Borden, MD; Research Institute, Cleveland Clinic Foundation (Cleveland, OH): R.M. Ransohoff, MD; Department of Microbiology, New York University Medical Center (New York): Jan T. Vilcek, MD.
Acknowledgments
Supported by the National Institutes of Health, NINDS R01-26321 and Biogen, Inc., Cambridge, MA.
- Received October 5, 1998.
- Accepted February 5, 1999.
References
Disputes & Debates: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.