http://ipsapp003.lwwonline.com/content/getfile/3836/4/14/fulltext.htm
J Neuroophthalmol 2001 December;21(4):296-301
Laura J. Balcer, MD, MSCE
From the Division of Neuro-Ophthalmology,
Departments of Neurology and Ophthalmology, University of Pennsylvania
School of Medicine, Philadelphia, Pennsylvania.
JOURNAL OF NEURO-OPHTHALMOLOGY 2001;21:296-301
The development of new and more sensitive clinical outcome measures for research in multiple sclerosis (MS) has been fueled by the development of effective therapies. As such, active arm comparison studies that require more sensitive clinical outcome measures are now commonplace. The Kurtzke Expanded Disability Status Scale (EDSS), the most widely used measure of neurologic impairment in MS, is particularly designed for classifying patients with respect to disease severity but has been criticized for its noninterval scaling, emphasis on ambulation status, relatively reduced sensitivity in the mid and upper ranges of scores, and absence of adequate cognitive and visual components. In response to perceived difficulties with the EDSS, the National Multiple Sclerosis Society Clinical Outcomes Assessment Task Force has developed the Multiple Sclerosis Functional Composite (MSFC). The MSFC includes three components that yield objective and quantitative results: 1) the timed 25-ft walk, 2) the nine-hole peg test, and 3) the 3-second paced auditory serial addition test. This scale has the advantages of continuous scoring with a composite Z score, standardized protocols, and high degrees of reliability and validity. Candidate visual function outcome measures for the MSFC, including the low-contrast Sloan letter chart, are currently under investigation. In addition to measures of neurologic impairment, health-related quality of life (HRQOL) measures have gained increasing importance as clinical trial outcome measures. The MS Quality of Life Inventory, a disease-specific HRQOL measure, has been developed to capture self-reported neurologic dysfunction and the impact of MS upon activities of daily living. MS clinical trials of the future, particularly active-arm comparison studies, will require more sensitive clinical outcome measures such as the MSFC. Measures of visual function and HRQOL should also be incorporated to capture the broad scope of neurologic impairment and disability in MS populations.
Clinical outcome measures are critical to the assessment of disease status and progression for clinical trials and other research in multiple sclerosis (MS). The development of new and more sensitive clinical outcome measures for MS research has been necessitated not only by characteristics of the disease process (such as variability of disease course among patients), but also by the development of effective therapies for MS. As such, placebo-controlled trials in MS are becoming, for the most part, impractical, and active-arm comparison studies, which require more sensitive clinical outcome measures, are now commonplace. In addition, several important aspects of neurologic function in MS, particularly cognition and vision, have been recognized as undermeasured by traditional MS clinical outcome measures (1–7).
The most commonly used scales in MS trials measure impairment (objective clinical signs and symptoms), disability (personal limitations on activities of daily living imposed by the impairment), or health-related quality of life (HRQOL—an individual's assessment of how a health problem and its treatment affect the ability to perform valued activities and roles) (1,7–11). HRQOL measures may be categorized as generic (instruments that are designed to measure all important aspects of HRQOL, not only those of a specific disease or condition), or specific (scales that focus on aspects of HRQOL that are specific to a disease, function, condition, or population of interest) (7–15). Disease-specific HRQOL measures are often used to supplement generic HRQOL instruments, or may include a generic scale as a core measure (7,8,13). Table 1 includes an abbreviated list of the most commonly encountered clinical outcome measures used in MS research and clinical trials, including follow-up studies from the Optic Neuritis Treatment Trial (16–46). Scales indicated by an asterisk in Table 1 are discussed in detail in this article.
TABLE 1. Commonly used clinical
outcome measures in MS research and clinical trials
KURTZKE EXPANDED DISABILITY STATUS
SCALE
The Kurtzke Expanded Disability Status Scale (EDSS) is the most widely used measure of neurologic impairment in MS clinical trials (27,47,48). The EDSS was derived from the Disability Status Scale (DSS) originally introduced by Kurtzke in 1955 (26). The DSS was an ordinal (noncontinuous) scale consisting of 10 steps or disease stages, primarily based on the clinical neurologic examination and patient ambulation status. In 1961, a set of functional groups (analogous to the functional systems (FS) that are used in the current EDSS scoring system) was added to complement the DSS, including categories for pyramidal (motor), cerebellar, brainstem, sensory, bowel/bladder, visual, mental, and other functions (49). FS scores were then used to generate DSS scores from 0 to 4. Scores above 4 reflected more advanced disease and were dependent on ambulation status (the patient's ability to walk for minimal distances with or without unilateral, bilateral, or wheelchair assistance).
The DSS thus formed the basis for the current EDSS proposed by Kurtzke in 1983 (27). EDSS ratings range from 0 to 10.0, with 0.5-unit increments (except between 0 and 1) rather than 1-unit increments as used in the DSS. For ratings of 4.0 or lower, the EDSS score is based on scores from eight FS, including pyramidal, cerebellar, brainstem, sensory, bowel and bladder, visual, and cerebral (mental) function. Scores above 4.0 are highly dependent on the patient's ambulation status—primarily the ability to walk certain distances and a dependence on assistive devices (27,28). Within this range of EDSS scores, other neurologic findings captured by the FS scores (such as arm, cognitive, and visual function) are not reflected in the overall EDSS score. Because EDSS scores above 4.0 are primarily based on ambulation status (a score of 6.0 indicates use of unilateral assistance for ambulation), the sensitivity of the EDSS for reflecting change in neurologic impairment is greatly reduced within this range. This may greatly limit the capacity of the EDSS to detect treatment effects in MS clinical trials, particularly in active-arm comparison studies in which the differences between groups may be small yet clinically significant (1–7).
The EDSS is an ordinal or noncontinuous scale for which the quantitative distances between scores are not well defined. Therefore, summary statistics such as mean and standard deviation, as would be used for interval (continuous) scales, may not be entirely appropriate for the reporting of EDSS scores. Mean and standard deviation are used commonly in MS clinical trials and natural history studies, although median and range may represent more appropriate summary statistics. Despite the many potential criticisms of the EDSS, including its noncontinuous scoring, emphasis on ambulation status, and potential difficulties with sensitivity and reliability, the EDSS remains the standard clinical measure of neurologic impairment used in all major natural history studies and treatment trials and remains a useful tool for classifying patients clinically with respect to disease severity (1–4,27,47,48).
The EDSS is administered by neurologists as a standardized measure of neurologic impairment in the setting of clinical trials, natural history/observational studies, and clinical practice. A standardized protocol, Neurostatus, was recently developed to reduce potential interrater variability in the EDSS administration and scoring (28). The Neurostatus scoring protocol has been modified slightly from Kurtzke's EDSS with respect to scoring within the 6.0–6.5 range (28). These changes are based on distance walked with unilateral versus bilateral assistance. A training CD-ROM for EDSS administration and scoring accompanies the Neurostatus system.
MULTIPLE SCLEROSIS FUNCTIONAL COMPOSITE MEASURE
In response to perceived difficulties with the EDSS, the Multiple Sclerosis Functional Composite (MSFC) has been recently developed by the National Multiple Sclerosis Society Clinical Outcomes Assessment Task Force (MS Task Force) (4–6). This new scale has three components that yield objective and quantitative results: 1) the timed 25-ft walk, 2) the nine-hole peg test (50), and 3) the 3-second paced auditory serial addition test (51). Unlike the EDSS, the MSFC may be feasibly administered by trained technicians or other nonphysician personnel. Testing time for the MSFC is brief (approximately 15 minutes), and the facilities required for testing are simple (quiet examination room with table/desk and hallway for timed 25-ft walk) (6,52).
Currently in use as the primary outcome measure for neurologic impairment in a recently completed randomized trial of interferon beta-1a in secondary progressive MS (IMPACT Study), the MSFC also has several advantages from a psychometric standpoint, including multidimensional design (measures several aspects of MS impairment), continuous scoring (composite Z score), standardized protocols, and high degrees of concurrent and predictive validity when compared with the EDSS and with surrogate markers for MS disease activity such as magnetic resonance imaging (4–6,52). In a study of participants in a phase III trial of interferon beta-1a for relapsing MS (52), MSFC scores correlated significantly with EDSS at baseline (rs = –0.42, P < 0.0001) and at 2 years (rs = –0.68, P < 0.0001). Baseline MSFC scores were also significantly predictive of magnetic resonance imaging brain parenchymal fraction (a measure of degree of brain atrophy/neuronal loss) at 2 years (rs = 0.52, P < 0.0001).
A single continuous score for the MSFC (referred to as the composite Z score) is derived by combining Z scores for each of the three components (4–6). A Z score represents the number of standard deviation units that a patient's score is above or below the average score from a standardized population (pooled MS population or clinical trial baseline data selected to represent the standard). Z scores, as used for the MSFC and its components, are continuous scores (mean and standard deviation may be appropriately used as summary statistics), are more sensitive to changes over time, and allow direct comparison of scores with standard MS or control populations (6). Administration of the MSFC by nonphysician personnel also permits more thorough masking of neurologists to patient outcome measure results during the course of MS clinical trials, thus reducing the potential for bias. Standardized protocols for administration of the MSFC have been developed by the MS Task Force (53).
VISUAL FUNCTION OUTCOME MEASURES
FOR MS RESEARCH AND CLINICAL TRIALS
Visual impairment is a leading cause
of symptoms in patients with MS (54–57). The quantitative assessment of
visual function in MS clinical trials (the EDSS), however, has been generally
limited to measures of Snellen visual acuity (58). As recognized by the
MS Task Force, the MSFC, despite its many advantages, does not yet include
an assessment of visual function (4–6). In the evaluation of candidate
MSFC visual components, Snellen acuity did not change over time or demonstrate
concurrent changes over time with EDSS scores (5). There are two potential
reasons for this: 1) Snellen acuity may not be sufficiently sensitive in
patients with MS and 2) the potential for visual acuity to influence overall
EDSS scores may be limited (the visual function score may become “buried”
within the overall score), particularly among patients who have difficulty
with ambulation (59). The need for evaluation of new candidate visual measures
for assessing clinical outcomes in patients with MS is clear.
Numerous investigations have indicated that measures of contrast sensitivity and contrast letter acuity may be the most sensitive measures of visual dysfunction in patients with MS, even among those with Snellen acuities of 20/20 or better (23,25,59–73). A new set of contrast letter acuity charts, the Low-Contrast Sloan Letter Charts (LCSLC), has demonstrated a high degree of interrater reliability when used for testing in patients with MS and in disease-free controls of similar age (59). Contrast letter acuity testing using the LCSLC also captures aspects of visual and neurologic function in MS that are not captured by Snellen visual acuity or ambulation status (59). More recent studies have shown that LCSLC testing is both a valid and feasible measure of visual function among patients with relapsing/remitting and secondary progressive MS (74). In a cross-sectional study of 50 patients with MS at the University of Pennsylvania, rank correlations of LCSLC scores (1.25% contrast level) with MSFC and EDSS scores were significant yet modest to moderate in magnitude (rs = 0.49, P = 0.0006 for LSCLC versus MSFC; rs = –0.37, P = 0.008 for LSCLC versus EDSS), supporting a potential role for LCSLC as an MSFC visual component (74).
MEASURES OF DISEASE-SPECIFIC AND VISION-SPECIFIC HRQOL IN MS
MS may have important and lasting effects on HRQOL that are not entirely captured by measures of neurologic impairment (7,9,10,40,41,58,75,76). HRQOL refers to an individual's assessment of how a health problem and its treatment affect the ability to perform valued activities and roles (7,11,12). HRQOL measures are unique in that they capture the patient's perspective on the impact of illness or treatment. There is consensus among MS investigators that HRQOL measures should be incorporated into ongoing data collection regarding the effects of MS and its treatments (7,8). Generic measures of HRQOL, including the 36-item Short-Form Health Survey from the Medical Outcomes Study (SF-36) (33–36), have demonstrated the impact of MS on important aspects of functioning and well-being. Although they demonstrate construct validity in MS, generic scales may not capture the impact of specific symptoms (7,8,40,41). The use of vision-specific HRQOL measures, including the National Eye Institute Visual Functioning Questionnaire (NEI-VFQ) and the 25-item National Eye Institute Visual Functioning Questionnaire (VFQ-25), in patients with MS has been recently investigated among participants in the Optic Neuritis Treatment Trial (21,22,42–45).
Multiple Sclerosis Quality of Life Inventory
The Multiple Sclerosis Quality of Life Inventory (MSQLI) has been developed as a disease-specific HRQOL instrument to complement measures of neurologic impairment (the EDSS and MSFC) in patients with MS (7,39). The MSQLI is a health profile measure that consists of a core generic scale, the SF-36, and nine other scales that capture various symptoms and aspects HRQOL that have been identified as important to patients with MS (7). Although the SF-36 captures major domains of general health, the nine symptom-specific subscales of the MSQLI capture fatigue, pain, sexual satisfaction, bladder function, bowel function, perceived visual function, perceived cognitive function, emotional status, and social functioning (Table 2) (7).
TABLE 2. VFQ-25 and MSQLI
subscales and summary measures
Fischer (7) recently reported high
levels of reliability and evidence supporting content and construct validity
for this measure in a large-scale field test. Several design features of
the MSQLI have led to its use as a measure of disease-specific HRQOL measure,
including multidimensional design (nine symptom-specific scales) and inclusion
of an established generic core (the SF-36). The reliability and validity
of the SF-36 have been established across a variety of conditions, and
population norms from patients with a variety of conditions, including
visual impairment, are available (33–36). Two summary scales for the SF-36,
the Physical Component Summary and the Mental Component Summary, are included
among the MSQLI subscales, as detailed in Table 2.
National Eye Institute Visual Functioning Questionnaires
Visual function and self-perceived visual impairment are important aspects of HRQOL in patients with MS (22,58). Accordingly, a five-item measure of self-perceived visual function, the Impact of Visual Impairment Scale, has been included as one of the nine symptom-specific subscales of the MSQLI (Table 2). Independent of the development of the MSQLI, the 25-item National Eye Institute Visual Function Questionnaire (VFQ-25) has been validated as a measure of vision-specific HRQOL (44,45). The VFQ-25, and the larger scale from which it was developed, the 51-item NEI-VFQ (42,43), are not specific to MS but have been used to demonstrate self-perceived visual impairment in patients with a variety of ocular disorders, including recovered optic neuritis, glaucoma, age-related macular degeneration, and many others. The VFQ-25 consists of 12 subscale scores and an overall composite score (45), as outlined in Table 2.
Signs and symptoms of visual impairment are strongly related to overall HRQOL, particularly among patients with MS and optic neuritis (58,77,78). In the Optic Neuritis Treatment Trial cohort, vision-specific HRQOL scales captured self-perceived visual dysfunction at 6 months and at 5 to 8 years after an attack of acute optic neuritis (21,22). In the latter study, the 51-item NEI-VFQ results were similar when items from the short-form version (VFQ-25) only were used for analysis (22). Self-perceived visual dysfunction in this study was more common among those patients from the Optic Neuritis Treatment Trial who had developed clinically definite MS. This relationship was observed even though neurologic impairment in these patients was generally mild (70% of the 134 patients with MS had EDSS scores 2.5, indicating minimal disability).
Self-reported Visual Dysfunction in MS
Studies among clinically heterogeneous cohorts of MS patients have provided evidence that the VFQ-25 captures aspects of visual dysfunction that are not entirely captured by visual acuity or ambulation status (46). Modest but significant correlations of VFQ-25 composite scores with binocular visual acuity (rs = 0.33, P = 0.003) support construct validity for VFQ-25 scores in MS populations (the VFQ-25 is measuring some aspect of visual dysfunction in MS) (79). Additional items, with content more specific to MS and other neuro-ophthalmologic disorders (such as double vision), may enhance the capacity of the VFQ-25 to capture self-perceived visual dysfunction in patients with MS. Further examination of how measures of visual function may relate to vision-and disease-specific HRQOL in MS is needed to determine which visual function outcome measures best capture aspects of HRQOL most valued by patients.
CONCLUSION
Clinical outcome measures are crucial
to the demonstration of disease progression and treatment efficacy in MS
research. With the advent of active-arm comparison studies of MS therapies,
the need for more sensitive and reliable yet practical measures of neurologic
impairment has led to the development of the MSFC. Despite the multidimensional
nature of the MSFC (it includes measures of arm, leg, and cognitive function),
measures of visual function such as the LCSLC must undergo further investigation
for potential inclusion. Given the increasing importance of assessing HRQOL
as a factor in patient outcomes, measures such as the MSQLI and VFQ-25
are likely to be included in future MS clinical trials. Despite the advantages
of the MSFC, the EDSS—the traditional clinical outcome measure for MS research—will
continue to be of importance for categorizing MS patients with respect
to neurologic impairment and disability.
REFERENCES
Copyright © 2001 Lippincott
Williams & Wilkins
Address correspondence and reprint requests
to Laura J. Balcer, MD, 3 E. Gates, 3400 Spruce Street, Philadelphia, PA
19104, USA;