`
`Multiple Sclerosis 2006; 12: 782 786
`
`How similar are commonly combined criteria for EDSS
`progression in multiple sclerosis?
`
`JJ Kragt1, JM Nielsen1, FAH van der Linden1,2, BMJ Uitdehaag1,3 and CH Polman1
`
`Introduction Measuring disease progression is an important aspect of multiple sclerosis (MS)
`clinical trials. Commonly applied disability endpoints include time to clinically meaningful Expanded
`Disability Status Scale (EDSS) change, or the number of patients in whom such a change has
`occurred. Typically, clinically meaningful EDSS change has been defined as a change of 1.0 point on
`Kurtzke’s EDSS in patients with an entry EDSS score of 5.5 or lower, or 0.5 point in patients with a
`higher EDSS score. Our goal was to evaluate whether these changes can be considered as similar.
`Therefore, we compared EDSS changes to corresponding changes in the Guy’s Neurological
`Disability Scale (GNDS), which is a measure of patient perceived disability, and the Multiple Sclerosis
`Functional Composite (MSFC), which is an examination-based quantitative scoring of neurological
`impairment.
`Methods
`From a large longitudinal database, we selected two groups of patients with a clinically
`meaningful change in EDSS score according to the usual criteria: patients with EDSS change ]/1.0
`for baseline EDSS 5/5.5 and patients with EDSS change ]/0.5 for baseline EDSS ]/6.0. We compared
`changes in GNDS sum score and in MSFC score between both groups.
`In the group with baseline EDSS ]/6.0, GNDS and MSFC changes were higher than in
`Results
`patients with baseline EDSS 5/5.5. The difference in change was 1.00 (95% confidence interval (CI):
`/0.35 to 2.36) for the GNDS and 0.412 (95% CI: 0.300 0.525) for the MSFC.
`Conclusion Our results indicate that a 0.5 point EDSS change in patients with baseline EDSS
`]/6.0 cannot be considered equal to a 1.0 point change in patients with baseline EDSS 5/5.5.
`Multiple Sclerosis 2006; 12: 782 786. http://msj.sagepub.com
`
`Key words: clinical scale; EDSS; GNDS; MSFC; multiple sclerosis
`
`Introduction
`
`Measurement of disability is indispensable in asses-
`sing the efficacy of experimental therapeutic agents
`in multiple sclerosis (MS). Clinical scales are being
`used as primary or secondary outcome measures for
`recording disease progression in clinical
`trials.
`Despite several methodological
`limitations,
`the
`Expanded Disability Status Scale (EDSS) [1], is still
`used as a gold standard for measuring impairment
`and disability in MS.
`Sensitivity to detect disease progression, also
`called responsiveness, is a key attribute for any
`assessment tool in MS clinical trials [2]. Due to its
`ordinal and non-continuous nature,
`the mean
`
`change in EDSS is an inappropriate endpoint [3],
`and therefore, a definition of treatment failure
`based on change in score from baseline has been
`introduced a change in EDSS beyond a certain
`cut-off that is considered to be relevant and sus-
`tained during two consecutive examinations or for
`a certain length of time [4].
`In previous trials [5,6], an EDSS change of 1.0
`point sustained for three or more months has been
`considered as clinically meaningful for patients
`with a baseline EDSS score of B/6.0. For patients
`with a higher EDSS score, a clinically meaningful
`change has been defined as a 0.5 point EDSS
`change. This guideline [7] has been defined as a
`logical consequence of the following characteristics
`
`1 Department of Neurology, VU University Medical Centre, Amsterdam, the Netherlands
`2 Department of Medical Psychology, VU University Medical Centre, Amsterdam, the Netherlands
`3 Department of Clinical Epidemiology and Biostatistics, VU University Medical Centre, Amsterdam, the Netherlands
`Author for correspondence: JJ Kragt, Department of Neurology, VU University Medical Centre, De Boelelaan 1117, PO
`Box 7057, 1007 MB Amsterdam, the Netherlands. E-mail: jj.kragt@vumc.nl
`Received 22 September 2005; accepted 24 February 2006
`
`– 2006 SAGE Publications
`
`10.1177/1352458506070931
`
`MYLAN INC. EXHIBIT NO. 1040 Page 1
`
`
`
`How similar are commonly combined criteria for EDSS progression in MS?
`
`783
`
`of EDSS: differing staying times at specific EDSS
`levels and varying reproducibility of EDSS through-
`out its range.
`First, the definition of a clinically meaningful
`change on EDSS should take into account the
`variable staying times at specific EDSS levels [8].
`The mean staying times proved to be greatest at
`Disability Status Scale (DSS; as EDSS was formerly
`called) 1 and 7 and least for DSS 4 and 5. Second,
`several studies have been performed to assess intra-
`and inter-rater agreement of the EDSS [9 12]. In
`these studies, greater variability was observed in the
`lower part of the scale, showing that agreement
`depends on baseline EDSS and on definitions of
`agreement expressed by difference in EDSS scores.
`In most clinical trials, patients showing a change
`of at least 1.0 point in the lower range of EDSS or at
`least 0.5 point in the upper EDSS range are com-
`bined and reported as a single number,
`thus
`implicitly assuming that these changes are equal
`or at least similar. The purpose of the present study
`was to evaluate whether indeed these changes can
`be considered similar. Therefore, we selected from a
`large longitudinal database those patients in whom
`such EDSS changes, according to one of both
`criteria, had occurred and compared, between the
`groups, corresponding changes in two external
`standards, the Guy’s Neurological Disability Scale
`(GNDS), a measure of patient perceived disability
`[13], and the Multiple Sclerosis Functional Compo-
`site (MSFC), an examination-based quantitative
`scoring of neurological impairment [14].
`
`Methods
`
`Patients and test procedures
`
`The database consisted of 662 patients with MS
`[15], who had undergone a series of test procedures
`as part of a health status assessment program
`designed to improve individual patient care at the
`MS Center of the VU University Medical Centre. No
`criteria for age, gender, disability level or MS
`subtype were applied during selection of data for
`analysis.
`Patients were first selected on the basis of the
`availability of repeated EDSS, GNDS and MSFC
`examinations, with a time interval of at least 265
`days. These examinations had been performed
`during the same visit under carefully standardized
`conditions by well-trained medical doctors, as
`described previously [16,17]. To standardize neuro-
`logical examination as much as possible, we made
`use of the Neurostatus (Version 2, CD-ROM). The
`GNDS, which is a patient-based interview that
`captures the major domains of disabilities in MS,
`contains 12 subcategories that were scored and
`
`summed to create the GNDS sum score, ranging
`from 0 to 60 [13].
`The MSFC consists of three quantitative mea-
`sures: the Timed 25-foot Walk (T25FW) to assess
`lower limb disability, the 9-hole Peg Test (9HPT), a
`measure of upper limb function, and the Paced
`Auditory Serial Addition Test (PASAT) which esti-
`mates cognitive disability. The quantitative results
`of the three tests are combined into a composite
`which makes it a sensitive instrument that is able to
`detect small clinical changes. For creating the MSFC
`score, Z -scores were calculated for the T25FW,
`9HPT and PASAT [14]. Z -scores were obtained using
`means and standard deviations of an external
`reference population, consisting of a wide range of
`MS patients [17]. As the Z -score sign had to be the
`same for all three tests, the mean of the 9HPT was
`transformed to its inverse before creating the Z -
`score and the Z -score of the T25FT was multiplied
`by 1. The composite score was calculated by adding
`the three Z -scores and dividing it by three: MSFC/
`(Z[1/(9-HPT), average]/Z[T25FT]/Z[PASAT])/3 [18]. Inabil-
`ity to perform a test of the MSFC due to MS-related
`symptoms, was scored with the maximum time
`allowed for the T25FT (180 seconds) and 9HPT
`(300 seconds) and with the worst score for the
`PASAT (0) [17,19].
`
`Analysis
`
`Results were analysed in several ways. First, we
`selected those patients with a predefined clinically
`meaningful change in EDSS score according to the
`usual definition: EDSS change ]/0.5 for baseline
`EDSS ]/6.0 (subset A) and EDSS change ]/1.0 for
`baseline EDSS 5/5.5 (subset B). After selection, we
`compared changes in GNDS sum score and MSFC
`score in subset A to those in subset B. Second, we
`further subdivided subset B in two groups according
`to baseline EDSS, which resulted in the formation
`of three disability strata: EDSS ]/6.0 (subset A),
`EDSS 4.0 5.5 (subset B1) and EDSS 0 3.5 (subset
`B2), and analysed GNDS and MSFC changes in
`these three groups. Finally, we explored newly
`defined EDSS changes in subset A and subset B
`that would give rise to more or less equal GNDS and
`MSFC changes.
`
`Statistics
`
`We evaluated differences in GNDS and MSFC
`changes between groups using Student’s t -tests.
`We calculated point estimates of these differences
`in change with corresponding 95% confidence
`intervals (CI). To correct for multiple comparisons,
`we considered P values B/0.01 as significant.
`
`http://msj.sagepub.com
`
`Multiple Sclerosis 2006; 12: 782 786
`
`MYLAN INC. EXHIBIT NO. 1040 Page 2
`
`
`
`784
`
`JJ Kragt et al.
`
`Table 1 Descriptives
`
`Total group
`
`Subset A
`(EDSS ]/6.0)
`
`Subset B
`(EDSS 5/5.5)
`
`Subset B1
`(EDSS 4.0 5.5)
`
`Subset B2
`(EDSS 5/3.5)
`
`n
`Female/male
`RR/SP/PP
`Age, years (SD)
`Follow-up duration, days (range)
`
`282
`180/102
`155/62/45
`41.9 (10.2)
`691 (268 1715)
`
`70
`37/33
`12/38/19
`45.2 (11.1)
`716 (296 1715)
`
`212
`143/69
`143/25/25
`40.8 (9.7)
`682 (268 1669)
`
`51
`30/21
`20/18/12
`46.8 (8.6)
`779 (308 1631)
`
`161
`113/48
`123/7/13
`39.0 (9.3)
`651 (268 1669)
`
`Results
`
`We selected 606 patients (from a database contain-
`ing 662 patients) who were examined twice with a
`time interval of at least 265 days. A total of 282
`patients
`experienced a
`clinically meaningful
`change in EDSS score, as defined previously. Of
`these patients, 102 (36%) were male and 180 (64%)
`female. Mean age at baseline was 41.9 years (stan-
`dard deviation (SD) 10.2). Most patients were
`diagnosed as having relapsing-remitting (RR) MS
`(55%), smaller proportions as having secondary
`progressive (SP) MS (22%) and primary progressive
`(PP) MS (16%) [20]. Average time from baseline
`to follow-up measurement was 691 days, range
`268 1715 days.
`Of these 282 patients, 70 were severely disabled
`(subset A). Mean follow-up duration in this sub-
`group was 716 days (range: 296 1715) and median
`EDSS change was 0.5 (range: 0.5 2.0). Mean GDNS
`sum score was 20.48 (SD: 6.78) at baseline and
`23.62 (SD: 7.00) at follow-up, resulting in a GNDS
`change of 3.14 (SD: 4.97). MSFC scores in this group
`were /0.611 (SD: 0.818) at baseline and /1.128
`(SD: 0.865) at follow-up. Therefore, MSFC change
`measured /0.517 (SD: 0.655). These calculations
`
`Table 2 GNDS and MSFC scores in the different subsets
`
`were also performed in the subgroup of 212 patients
`who were mildly or moderately disabled (subset B).
`This subgroup experienced a median EDSS change
`of 1.5 (range: 1.0 4.0) after an average follow-up
`duration of 682 days (range: 268 1669). Mean
`GNDS
`sum score
`in this group was 10.83
`(SD: 6.32) at baseline and 12.97 (SD: 7.05) at
`follow-up, which resulted in a GNDS change of
`2.14 (SD: 4.96). MSFC score at baseline was 0.462
`(SD: 0.435) and at follow-up 0.357 (SD: 0.494),
`which gave rise to a MSFC change of /0.105
`(SD: 0.251). When comparing GNDS changes in
`both groups, a non-significant trend towards a
`higher GNDS change in subset A was observed;
`the difference in change was 1.00 (95% CI: /0.35
`to 2.36). Regarding the MSFC, we found a difference
`in change of 0.412 (95% CI: 0.300 0.525).
`The exact same analyses were performed after
`subdividing subset B in subsets B1 and B2. Differ-
`ences in changes between subsets B1 and B2 were
`not significantly different, neither for GNDS nor for
`MSFC. Descriptives and scores on GNDS and MSFC
`are shown in more detail in Tables 1 and 2.
`Finally, we investigated different cut-off values
`for EDSS changes that would result in comparable
`changes in GNDS and MSFC in subsets A and B. For
`this, we used the change in subset A as reference
`
`Subset A
`(EDSS ]/6.0)
`
`0.5 (0.5 2.0)
`
`20.48 (6.78)
`
`23.62 (7.00)
`
`3.14 (4.97)
`
`/0.611 (0.818)
`
`/1.128 (0.865)
`
`/0.517 (0.655)
`
`Median change in
`EDSS (range)
`Baseline GNDS
`(SD)
`Follow-up GNDS
`(SD)
`Change in GNDS
`(SD)
`Difference in GNDS
`change (95% CI)
`Baseline MSFC
`(SD)
`Follow-up MSFC
`(SD)
`Change in MSFC
`(SD)
`Difference in MSFC
`change (95% CI)
`
`A versus B
`
`Subset B
`(EDSS 5/5.5)
`
`Subset B1
`(EDSS 4.0 5.5)
`
`B1 versus B2
`
`Subset B2
`(EDSS 5/3.5)
`
`1.5 (1.0 4.0)
`
`1.5 (1.0 3.5)
`
`1.5 (1.0 4.0)
`
`10.83 (6.32)
`
`16.63 (5.78)
`
`12.97 (7.05)
`
`18.94 (6.49)
`
`2.14 (4.96)
`
`2.31 (5.33)
`
`1.00
`(/0.35 to 2.36)
`
`0.23
`(/1.80 to 1.35)
`
`0.462 (0.435)
`
`0.096 (0.436)
`
`0.357 (0.494) /0.010 (0.541)
`
`/0.105 (0.251) /0.107 (0.298)
`
`0.412
`(0.300 0.525)
`
`0.003
`(/0.083 to 0.088)
`
`8.98 (5.29)
`
`11.07 (6.10)
`
`2.09 (4.86)
`
`0.570 (0.372)
`
`0.466 (0.422)
`
`/0.104 (0.237)
`
`EDSS, Expanded Disability Status Scale; MSFC, Multiple Sclerosis Functional Composite; GNDS, Guy’s Neurological Disability Scale.
`
`Multiple Sclerosis 2006; 12: 782 786
`
`http://msj.sagepub.com
`
`MYLAN INC. EXHIBIT NO. 1040 Page 3
`
`
`
`How similar are commonly combined criteria for EDSS progression in MS?
`
`785
`
`category, since 0.5 is the minimal detectable EDSS
`change. Results are shown in Table 3. An EDSS
`change of ]/2.0 in patients with EDSS 5/5.5
`resulted in a GNDS change that was roughly the
`same as that in patients with EDSS ]/6.0 who
`experienced an EDSS change of ]/0.5 (GNDS
`change 4.00 versus 3.14). To level out MSFC
`changes, a higher EDSS change was needed: ]/2.5
`in patients with EDSS 5/5.5 was more or less equal
`to an EDSS change ]/0.5 in patients with EDSS
`]/6.0 (MSFC change /0.407 versus /0.517).
`When evaluating different cut-off values for EDSS
`changes in the three subsets A, B1 and B2, similar
`results were obtained (data not shown).
`
`Discussion
`
`The exact definition of a clinically meaningful
`change in EDSS score is of great
`importance,
`especially in MS clinical trials. In most trials, this
`change has been defined as either a 1.0 point
`change for patients with an EDSS score at study
`entry of 5.5 or lower or a change of 0.5 point for
`more disabled patients (EDSS ]/6.0) and statistical
`analysis plans typically combine these patient
`groups. Nonetheless, the assumption that these
`two changes have equal clinical impact has never
`been properly examined. By using other clinical
`measurements, both subjective and objective, we
`were able to compare the changes associated with a
`1.0 point and a 0.5 point change on EDSS, respec-
`tively, in patients with varying degrees of disability.
`We found that concomitant GNDS and MSFC
`changes were considerably higher in patients with
`severe disability (EDSS ]/6.0) who experienced an
`EDSS change ]/0.5 point compared to GNDS and
`MSFC changes in patients with mild and moderate
`disability who had at least 1.0 point EDSS change.
`In order to compare the aforementioned EDSS
`changes, we used two other clinical measurements.
`First, we assessed the subjective clinical impact of
`EDSS changes by analysing GNDS changes. This
`instrument has been highly estimated as a measure-
`ment to assess patient perceived disability in differ-
`ent
`functional domains of MS. Second, we
`compared these EDSS changes in a more quantita-
`tive and objective manner by making use of the
`
`MSFC. Consequently, both external scales provid-
`ing similar results, we show that these predefined
`EDSS changes cannot be considered as equal.
`Obviously, the two external measurements have
`their limitations. Because the GNDS is a more
`subjective measurement that incorporates the pa-
`tient’s perspective, dissimilarities between GNDS
`and EDSS changes are a natural consequence. This
`can be illustrated by the fact that a number of
`patients experienced a clinically meaningful EDSS
`worsening, but did not deteriorate on GNDS (28%
`of patients in subset A, 37% of subset B). Concern-
`ing MSFC, the clinical meaningfulness of MSFC
`changes still needs to be clarified. Since this instru-
`ment has only recently been implemented in
`clinical trials, further research is warranted on this
`topic. Moreover, an explanation for the larger
`MSFC changes found in patients with severe dis-
`ability could lie in the fact that these patients were
`unable to perform one or more of the tests at their
`follow-up visit and, therefore, were assigned the
`maximum time or worst score (eg, 0 for the PASAT).
`Finally, MSFC and EDSS possibly measure different
`dimensions of disability. This can be demonstrated
`by the finding that a number of patients who
`worsened on EDSS did not worsen on MSFC (23%
`in subset A, 32% in subset B).
`Another pitfall of this study was the fact that
`EDSS changes were not confirmed in any manner.
`In clinical trials, a sustained EDSS change is con-
`firmed by repeated measurements after three or six
`months. In our study this was not the case.
`When using GNDS and MSFC changes from our
`study to titrate EDSS changes to have a comparable
`impact on patients with severe versus mild or
`moderate disability, we found that a 0.5 point
`EDSS change in patients with a score of 6.0 or
`higher more or less corresponds to a 2.0 or 2.5 point
`change in patients with a score of 5.5 or lower. We
`found no evidence that patients with EDSS of 5.5 or
`lower should further be separated in mild versus
`moderate disability in order to fine-tune the impact
`of EDSS changes.
`These results support the need for careful recon-
`sideration of the criteria for clinically meaningful
`EDSS changes, and the desirability of describing
`both groups separately in clinical trial reports.
`
`Table 3 Different cut-off values for EDSS changes and associated GNDS and MSFC changes
`
`Baseline EDSS score range
`
`Cut-off EDSS change
`
`Associated GNDS change
`
`Associated MSFC change
`
`]/6.0
`5/5.5
`
`0.5
`1.0
`1.5
`2.0
`2.5
`3.0
`
`3.14
`2.14
`2.69
`4.00
`7.36
`7.88
`
`/0.517
`/0.105
`/0.117
`/0.144
`/0.407
`/0.501
`
`http://msj.sagepub.com
`
`Multiple Sclerosis 2006; 12: 782 786
`
`MYLAN INC. EXHIBIT NO. 1040 Page 4
`
`
`
`786
`
`JJ Kragt et al.
`
`Acknowledgements
`
`The MS Centre of the VU University Medical Centre
`is partially funded by a program grant of the Dutch
`MS Research Foundation. No competing interests
`were declared.
`
`References
`
`1. Kurtzke JF. Rating neurologic impairment in multiple
`sclerosis: an expanded disability status scale (EDSS).
`Neurology 1983; 33: 1444 52.
`2. Rudick R, Antel J, Confavreux C, Cutter G, Ellison
`G, Fischer J et al . Clinical outcomes assessment in
`multiple sclerosis. Ann Neurol 1996; 40: 469 79.
`3. Wingerchuk DM, Noseworthy JH, Weinshenker
`BG. Clinical outcome measures and rating scales in
`multiple sclerosis trials. Mayo Clin Proc 1997; 72: 1070
`79.
`4. Weinshenker BG, Issa M, Baskerville J. Meta-analy-
`sis of the placebo-treated groups in clinical trials of
`progressive MS. Neurology 1996; 46: 1613 19.
`5. European Study Group on interferon beta-1b in
`secondary progressive MS. Placebo-controlled multi-
`centre randomised trial of interferon beta-1b in treatment
`of secondary progressive multiple sclerosis. Lancet 1998;
`352: 1491 97.
`6. Randomized controlled trial of interferon- beta-1a in
`results. Neurology
`secondary progressive MS: clinical
`2001; 56: 1496 504.
`7. Goodkin DE. EDSS reliability. Neurology 1991; 41: 332.
`8. Weinshenker BG, Rice GP, Noseworthy JH,
`Carriere W, Baskerville J, Ebers GC. The natural
`history of multiple sclerosis: a geographically based
`study. 4. Applications to planning and interpretation of
`clinical therapeutic trials. Brain 1991; 114: 1057 67.
`9. Amato MP, Fratiglioni L, Groppi C, Siracusa G,
`Amaducci L. Interrater reliability in assessing functional
`systems and disability on the Kurtzke scale in multiple
`sclerosis. Arch Neurol 1988; 45: 746 48.
`10. Francis DA, Bain P, Swan AV, Hughes RA. An
`assessment of disability rating scales used in multiple
`sclerosis. Arch Neurol 1991; 48: 299 301.
`
`11. Goodkin DE, Cookfair D, Wende K, Bourdette D,
`Pullicino P, Scherokman B et al . Inter- and intrarater
`scoring agreement using grades 1.0 to 3.5 of the Kurtzke
`Expanded Disability Status Scale (EDSS). Multiple Sclero-
`sis Collaborative Research Group. Neurology 1992; 42:
`859 63.
`12. Noseworthy JH, Vandervoort MK, Wong CJ, Ebers
`GC. Interrater variability with the Expanded Disability
`Status Scale (EDSS) and Functional Systems (FS) in a
`multiple sclerosis clinical trial. The Canadian Coopera-
`tion MS Study Group. Neurology 1990; 40: 971 75.
`13. Sharrack B, Hughes RA. The Guy’s Neurological
`Disability Scale (GNDS): a new disability measure for
`multiple sclerosis. Mult Scler 1999; 5: 223 33.
`14. Cutter GR, Baier ML, Rudick RA, Cookfair DL,
`Fischer JS, Petkau J et al . Development of a multiple
`sclerosis functional composite as a clinical trial outcome
`measure. Brain 1999; 122: 871 82.
`15. Poser CM, Paty DW, Scheinberg L, McDonald WI,
`Davis FA, Ebers GC et al . New diagnostic criteria for
`multiple sclerosis: guidelines for research protocols. Ann
`Neurol 1983; 13: 227 31.
`16. Hoogervorst EL, van Winsen LM, Eikelenboom
`MJ, Kalkers NF, Uitdehaag BM, Polman CH.
`Comparisons of patient self-report, neurologic examina-
`tion, and functional impairment in MS. Neurology 2001;
`56: 934 37.
`17. Kalkers NF, De Groof V, Lazeron RH, Killestein J,
`Ader HJ, Barkhof F et al . MS functional composite:
`relation to disease phenotype and disability strata.
`Neurology 2000; 54: 1233 39.
`18. Fischer JS, Jak AJ, Kniker JE, Rudick RA, Cutter G.
`Administration and scoring manual for the Multiple Sclerosis
`Functional Composite (MSFC) . New York, 2001.
`19. Hoogervorst EL, Kalkers NF, Cutter GR, Uitdehaag
`BM, Polman CH. The patient’s perception of a (reliable)
`change in the Multiple Sclerosis Functional Composite.
`Mult Scler 2004; 10: 55 60.
`20. Lublin FD, Reingold SC. Defining the clinical course of
`multiple sclerosis: results of an international survey.
`National Multiple Sclerosis Society (USA) Advisory Com-
`mittee on Clinical Trials of New Agents in Multiple
`Sclerosis. Neurology 1996; 46: 907 11.
`
`Multiple Sclerosis 2006; 12: 782 786
`
`http://msj.sagepub.com
`
`MYLAN INC. EXHIBIT NO. 1040 Page 5
`
`