You are here
Relationship between symptom change, relapse activity and disability progression in multiple sclerosis
Journal of the Neurological Sciences, Volume 362, 15 March 2016, Pages 121–126
Symptom changes may serve as a risk factor for relapse activity (RA) and disability progression (DP), which could facilitate multiple sclerosis (MS) treatment decisions.
To assess the relationship of symptom change with RA and DP.
We evaluated the relationship of symptom change with subsequent RA and DP using NARCOMS registry data reported over a five-year period. Symptom change was evaluated using both symptom worsening (SW) and average of Performance Scales (APS) scores. Disability progression was defined as a one-point or more increase in Patient-Determined Disease Steps (PDDS) score between two consecutive updates. Repeated measures logistic regression was used to investigate the relationship between symptom change and RA and DP.
SW and APS were both significant predictors of subsequent RA and DP. Both SW and APS have a significant interaction with levels of disability (Mildly Impaired versus Highly Impaired) for the prediction of the subsequent RA or DP. For Mildly Impaired MS subjects, both SW and APS were significant predictors of both RA and DP. However, for Highly Impaired MS subjects, SW did not significantly predict future RA and neither SW nor APS predicted disability progression.
Changes in self-reported overall symptomatology may precede and predict clinical relapse and future disability progression. The predictive power of symptom changes may only be present at lower levels of disability.
- Symptom worsening may be a prodromal event to a subsequent relapse or progression.
- Symptom worsening's predictive power may be limited to lower levels of disability.
- Patient reported outcomes may provide cues for earlier interventions by clinicians.
Keywords: Multiple sclerosis, Disease modifying therapy, Relapse activity, Disability progression, Symptom worsening, Patient-reported outcomes, PDDS.
Multiple Sclerosis (MS) is an inflammatory, immune-mediated disease that targets the central nervous system (CNS) , , and . MS affects approximately 400,000 individuals in the United States , , , and . Approximately 85% of these individuals have relapsing forms of MS (RRMS), where worsening most commonly occurs by relapse and incomplete recovery from acute relapses . In progressive forms of MS, worsening most commonly occurs by gradual progression of neurological disability in the absence of acute relapses  and . Even during remissions in RRMS, MRI studies show that CNS inflammatory activity may persist in the absence of acute symptom relapses . If left untreated, within 25 years approximately 90% of persons with RRMS will enter the secondary progressive phase of MS (SPMS) , , , , , and . In SPMS, disability progression is more resistant to treatment than in RRMS , , , , , , and . Therefore, preventing or delaying the accumulation of irreversible disability and the transition to SPMS are central goals of the disease management in patients with RRMS.
Relapse activity is often considered an indication of breakthrough disease and suboptimal response to disease-modifying therapy (DMT) in MS. Ongoing CNS inflammation involves both axonal damage and reparative processes. As more and more irreversible neurological damage occurs during the CNS inflammation, relapses eventually lead to accumulation of disability and disability progression (DP). Symptom change, with or without an acute relapse, could be a more subtle clinical manifestation of these inflammatory processes. Therefore, we hypothesized that symptom changes may be related to ongoing inflammation and serve as a risk factor for more overt clinical relapse activity and disability progression. A better understanding of the nature of such associations may provide useful information for disease management and treatment decisions. We evaluated this hypothesis by utilizing NARCOMS, a large, longitudinal MS patient registry where overall symptoms change and clinical relapses are reported every six months.
2. Methods and materials
2.1. Cohort disposition
This was a retrospective cohort study using data from the North American Research Committee on Multiple Sclerosis (NARCOMS) self-report registry from 2006 through 2010. Participants in this analysis reported having relapsing disease course in 2006 and completed two updates in each of the next five years yielding a total of 10 semi-annual updates per participant. Excluding those with missing surveys and using only those with complete data provided an equal space epoch of time for consistency and minimization of recall bias due to extended time between follow-up surveys. To assess the generalizability of the participants included in the analysis, we compared the demographics of these participants to those with less frequent survey responses; the two subsets of participants did not significantly differ on any sociodemographic characteristics.
2.2. Primary outcomes
The primary outcomes of the analyses were relapse activity and disability progression. Relapse activity was treated as a dichotomous variable based on participants' responses to a survey question as to whether or not the participant had a relapse in the past 6 months. NARCOMS defines a relapse as a “development of new symptoms or worsening of old symptoms that lasts longer than 48 h” that occurs at least 30 days after a previous relapse. In this study, all participants were given this definition at each survey prior to the questions being presented on the form. Disability status is reported using the Patient Determined Disease Steps (PDDS) . PDDS is a validated nine-point self-report instrument (ranging from 0 = normal to 8 = bedridden) which has been shown to correlate strongly with the Expanded Disability Status Scale . Overall disability progression (DP) was defined as at least a one-point increase in PDDS score between two consecutive updates.
2.3. Symptom change
Symptom change was evaluated using both symptom worsening and the Performance Scales scores. At each semi-annual survey, subjects were asked “Over the last 6 months, have your MS symptoms worsened in a gradual, progressive manner (not due to relapses or exacerbations)?” The responses to this question (Yes, No) were used as a measure of symptom worsening (SW).
Performance Scales (PS©)1 were used to assess specific symptoms in participants  and . The PS are a single question for each of eight domains: mobility, bowel/bladder, fatigue, sensory, vision, cognition, spasticity, and hand function. All of the subscales are scored using a six-point scale (0 = normal, 1 = minimal, 2 = mild, 3 = moderate, 4 = severe, or 5 = total disability), except the mobility subscale which includes a seventh response category (0 = normal, 1 = minimal gait, 2 = mild gait, 3 = occasional use of cane or unilateral support, 4 = frequent cane use, or 5 = severe gait, bilateral support, 6 = total gait disability or bedridden). An average Performance Scales score (APS) was determined for each update survey by taking the sum of the subscale scores (bowel/bladder, fatigue, sensory, vision, cognition, spasticity, and hand function) and dividing it by the number of PS subscales the participant completed. Mobility was not included in the APS as it was found to be highly correlated with PDDS (Spearman correlation coefficient = 0.96; p < 0.001). To ensure that APS scores were representative of the overall symptom severity of patients, participants had to have completed at least five out of seven Performance Scales (PS); 61 participants (0.2%) were excluded from the analysis due to incomplete PS data.
The start of the assessment period in 2006 is considered the baseline for socio-demographic covariates considered: age, gender, race, education, employment, insurance and household income level; and disease status including: PDDS and disease modifying treatment (DMT) status. Disease history is measured from NARCOMS enrollment information to 2006 including: age at symptom onset, year of symptom onset, year of diagnosis, disease duration, and relapse history.
2.4. Statistical analysis
These data permitted multiple intervals of assessment per participant and thus we used generalized estimating equations (GEE) to handle the dependence between the repeated measures. We used logistic regression models to evaluate the relationship between symptom progression and relapse activity in the subsequent 6 months and 12 months. Logistic regression models were also used to investigate the effect of symptom progression on disability progression in the subsequent 6 months. Additional covariates were chosen for the model using a forward selection procedure; in order to be included in the final model, covariates had to either be statistically significant at the α = 0.05 level or change the estimate of the effect for SW by more than 10%. We then repeated this model selection and fitting using disability progression as the response variable, instead of relapse activity.
3.1. Cohort characteristics
A total of 2605 participants with relapsing disease completed ten consecutive update surveys from 2006 to 2010. Of those, 1687 (65%) participants had PDDS levels ≤ 4 at baseline, considered to indicate mild impairment, or less severe, relapsing MS (Mildly Impaired MS cohort). We designated the 918 (35%) participants with more severe relapsing MS (i.e., PDDS levels > 4 at baseline) as the Highly Impaired MS cohort. The participant socio-demographics and clinical characteristics at baseline are presented in Table 1 and Table 2, respectively. The Mildly Impaired and Highly Impaired MS cohorts significantly differed in age, disease and symptom duration, gender ratio, education level, and use of disease-modifying treatment. The two cohorts did not significantly differ at baseline in rates of relapse activity or symptom worsening in the prior 6 months (p > 0.10 from a chi-square test for both). Overall, 1182 (45.4%) participants reported SW in the 6 months prior to the baseline 2006 survey. To obtain a more nuanced picture of symptom worsening, participants were asked the follow-up question “Compare your overall MS symptoms now with what you experienced 6 months ago (much worse, worse, a little worse, no change, a little better, better, or much better)”. Of the 1182 participants reporting symptom worsening, 58 (4.9%) reported that their symptoms were “much worse” than 6 months ago, 298 (25.2%) reported “worse” symptoms, and 760 (64.3%) responded that their symptoms were “a little worse” than 6 months prior; for the remaining 66 participants, despite having reported that their MS symptoms had worsened in the previous 6 months, responded to this question with “no change”, “a little better” or “better”.
|All MS (n = 2605)||Mildly impaired MS (n = 1687)||Highly impaired MS (n = 918)||p-Value MS type|
|Age (years): mean (SD)||52.4 (9.8)||50.4 (9.5)||56.1 (9.2)||< 0.0001|
|Disease duration (years): mean (SD)||14.5 (9.1)||12.4 (8.2)||18.4 (9.3)||< 0.0001|
|Symptom duration (years): mean (SD)||22.4 (11.5)||20.1 (11.2)||26.6 (11.0)||< 0.0001|
|Female: n (%n)||2000 (76.8)||1357 (80.4)||643 (70.0)||< 0.0001|
|Bachelor's degree or higher: n (%n)||1354 (52.3)||943 (56.2)||411 (45.0)||< 0.0001|
|On DMT: n (%n)||1861 (71.4)||1285 (76.2)||576 (62.8)||< 0.0001|
|All MS (n = 2605)||Mildly impaired MS (n = 1687)||Highly impaired MS (n = 918)||p-Value MS type|
|PDDS: median (range)||4 (0, 8)||2 (0, 4)||6 (5, 8)||NA|
|PDDS: n (%n)|
|0: normal||380 (14.6)||380 (22.5)||NA||NA|
|1: mild disability||361 (13.9)||361 (21.4)||NA||NA|
|2: moderate disability||210 (8.1)||210 (12.5)||NA||NA|
|3: gait disability||322 (12.4)||322 (19.1)||NA||NA|
|4: early cane||414 (15.9)||414 (24.5)||NA||NA|
|5: late cane||303 (11.6)||NA||303 (33.0)||NA|
|6: bilateral support||291 (11.2)||NA||291 (31.7)||NA|
|7: wheelchair/scooter||315 (12.1)||NA||315 (34.3)||NA|
|8: bedridden||9 (0.4)||NA||9 (1.0)||NA|
|Relapse activity in prior 6 months: n (%n)|
|Yes||934 (35.9)||599 (35.5)||335 (36.5)||0.622|
|No||1451 (55.7)||950 (56.4)||501 (54.6)|
|Unsure||219 (8.4)||137 (8.1)||82 (8.9)|
|Not answered||1 (0.04)||1 (0.06)||0 (0.0)|
|Symptoms Worseninga: n (%n)|
|Yes||1182 (45.4)||749 (44.4)||433 (47.2)||0.130|
|No||1409 (54.1)||933 (55.3)||476 (51.9)|
|Not answered||14 (0.5)||5 (0.3)||9 (1.0)|
a Response to question on semi-annual survey: “Over the last 6 months, have your MS symptoms worsened in a gradual, progressive manner (not due to relapses or exacerbations)?”
Overall, 934 (35.9%) participants reported a relapse in that same time period. Between Spring 2006 and Fall 2010, amongst those reporting SW in the preceding 6 months, the proportion that reported at least one relapse over the following 6 months decreased from 41.3% in the first survey interval to 23.4% over the successive 6 month intervals during this 5 year period of observation. Similarly, reported relapses amongst those without SW decreased from 31.1% to 15.2% over the same period. A similar trend was seen in each PS domain.
3.2. Symptom change and relapse activity
Age, baseline relapse activity, time period, and cohort (Mild Impaired vs Highly Impaired MS) were selected as additional covariates for the final logistic regression model for the effect of SW on subsequent relapse activity. SW was significantly associated with relapse activity both 6 months (OR = 1.22; p < 0.0001; 95% CI: 1.13 to 1.31) and 12 months later (OR = 1.26; p < 0.0001; 95% CI: 1.16 to 1.36) (Table 3). An interaction between MS severity and SW was added to the model to investigate whether the effect of SW on RA varied by MS severity; for both the 6 month and 12 month models, this interaction was statistically significant (results not shown). To further investigate this relationship, we fit the regression model to the Mildly and Highly Impaired MS cohorts separately. For the Mildly Impaired MS cohort, SW remained a significant predictor of relapse activity (p < 0.0001 for both the 6 and 12 month models). However, for the Highly Impaired MS cohort, SW was not significantly associated subsequent relapse activity (Table 3).
|Outcome||Predictor||All MS||Mildly impaired MS||Highly impaired MS|
|Odds ratio (95% CI)||p-Value||Odds ratio (95% CI)||p-Value||Odds ratio (95% CI)||p-Value|
|RA (6 months)a||SW||1.22 (1.13, 1.31)||< 0.0001||1.29 (1.18, 1.41)||< 0.0001||1.10 (0.98, 1.24)||0.113|
|RA (12 months)a||SW||1.26 (1.16, 1.36)||< 0.0001||1.35 (1.22, 1.49)||< 0.0001||1.10 (0.96, 1.25)||0.170|
|RA (6 months)b||APS||1.27 (1.20, 1.33)||< 0.0001||1.37 (1.28, 1.46)||< 0.0001||1.13 (1.05, 1.23)||0.002|
|RA (12 months)b||APS||1.26 (1.20, 1.33)||< 0.0001||1.33 (1.24, 1.42)||< 0.0001||1.18 (1.08, 1.28)||0.0002|
a Adjusted for age, baseline relapse activity, time period and cohort.
b Adjusted for age, baseline relapse activity, time period and cohort.
The same set of covariates (age, baseline relapse activity, time period, and cohort) was selected as the additional covariates for the final logistic regression model for the effect of APS on subsequent relapse activity. APS was significantly associated with relapse activity both 6 months (OR = 1.27; p < 0.0001; 95% CI: 1.20 to 1.33) and 12 months later (OR = 1.26; p < 0.0001; 95% CI: 1.20 to 1.33) (Table 3). An interaction between MS severity and APS was added to the model and was statistically significant for both the 6 month and 12 month models. When the two cohorts were analyzed separately, APS remained a statistically significant predictor of later relapse activity for both (Table 3).
3.3. Symptom change and disability progression
Disease modifying treatment, time period and cohort were selected as additional covariates for the final logistic regression model for the effect of SW on subsequent disability progression (i.e., an increase of PDDS score between two consecutive updates). SW was significantly associated with disability progression (OR = 1.13; p = 0.001; 95% CI: 1.05 to 1.22) (Table 4). There was a significant interaction between MS severity and SW (results not shown). An interaction between MS severity and SW was added to the model to investigate whether the effect of SW on RA varied by MS severity; for both the 6 month and 12 month models, this interaction was statistically significant (results not shown). As with the models for relapse activity, the Mildly Impaired and Highly Impaired MS cohorts were analyzed separately to further assess this interaction. For the Mildly Impaired MS cohort, SW remained a significant predictor of disability progression (p < 0.0001). However, for the Highly Impaired MS cohort, SW was not significantly associated with a later increase in PDDS score (p = 0.978) (Table 4).
|Outcome||Predictor||All MS||Mildly impaired MS||Highly impaired MS|
|Odds ratio (95% CI)||p-Value||Odds ratio (95% CI)||p-Value||Odds ratio (95% CI)||p-Value|
|DPa||SW||1.13 (1.05, 1.22)||0.001||1.22 (1.11, 1.33)||< 0.0001||1.00 (0.87, 1.16)||0.978|
|DPb||APS||1.28 (1.22, 1.34)||< 0.0001||1.40 (1.32, 1.48)||< 0.0001||1.06 (0.97, 1.16)||0.204|
a Adjusted for disease modifying treatment, time period and cohort.
b Adjusted for disease modifying treatment, health insurance, time period and cohort.
Disease modifying treatment, health insurance, time period and cohort were selected as additional covariates for the final logistic regression model for the effect of APS on subsequent disability progression (i.e., an increase of PDDS score between two consecutive updates). APS was significantly associated with disability progression (OR = 1.28; p < 0.0001; 95% CI: 1.22 to 1.34) (Table 4). There was a significant interaction between MS severity and APS (results not shown). When the two cohorts were analyzed separately, APS was a significant predictor of DP for Mildly Impaired MS patients (p < 0.0001), but not for Highly Impaired MS patients (p = 0.204).
This analysis of a longitudinal patient registry demonstrated that both SW and APS are significantly associated with relapse activity as well as disability progression. Furthermore, SW and APS were significant predictors of relapse and patient-reported disability progression in the Mildly Impaired MS patients. We did not observe a significant association between SW and relapse activity nor did we find evidence of significant associations between either SW or APS and disability progression in Highly Impaired MS patients.
These results may be because SW and APS measure not only irreversible disability progression but also the transient and reversible consequences of a recent relapse occurring at or shortly before the time of disability assessment . The worsening of symptoms resulting from a relapse can be a transient, fully reversible effect, but can also contribute to the accumulation of irreversible disability . In this study, participants were asked specifically to report symptom worsening that was “not due to relapses or exacerbations”. Accordingly, we found that relapses did not predict future symptom worsening (data not shown). This result suggests that participants are indeed not reporting symptom worsening because of residual deficits from a previous relapse and that the worsening of symptoms occurs prior to a relapse, otherwise the correlations would be the same since the intervals were staggered by 6 months. Therefore, symptom worsening may be an indicator for ongoing subtle inflammation and potentially predicting a relapse. This observation is consistent with the findings by Morgan et.al. which showed the occurrence of MRI activity measured by gadolinium enhanced lesions has predictive power for subsequent relapses which diminished as the time from the MRI increases , implying that symptom change is more predictive of relapse activity in the near future than the more distant future.
The notable decline in relapses over time could be due to a number of factors. One possibility is that as these participants transition from RRMS to SPMS, a drop in relapses could be occurring lowering the mean number of relapses over this extended observation period. In addition, the effects of aging may be reducing the number of relapses in concert with the disease processes or other age related processes. However, this year to year drop in relapse activity also parallels that seen in clinical trials over the years of follow-up .
There are limitations to the current study. First, since NARCOMS is a voluntary registry, the generalizability of these findings relies upon whether participants are representative of the MS population. Nevertheless, NARCOMS participants have previously been showed to be similar to MS populations described in other studies . Second, as these are self-reported data, the accuracy of the data depends on what patients recalled and reported in the survey. Participants may have different recall patterns and biases regarding symptom worsening versus relapses, that is, different telescoping frames. However, the study required two semi-annual surveys in each of the five years, which should have minimized this potential recall bias due to standardizing the time frames. Further, the SW more often preceded the relapse and is thus less likely a recall bias. The prospective associations should be relatively free of this form of bias as there are repeated surveys capturing overlapping epochs of time.
It is also important to point out the limitations of our study arising from the use of patient-reported outcomes. The accuracy of self-reported data depends heavily on patients' understanding the questions and terminologies used in the survey. To accurately answer an item, patients must understand the concepts and accurately recall the SW and relapse which occurred within an appropriate frame of reference. In addition, they must gauge their actual, rather than anticipated, symptom change. The specificity of these relatively complex tasks suggests that they may be disproportionately influenced by patient characteristics that impact the accuracy of the patient-reported outcomes . In order to mitigate any potential confusion and prevent the inconsistency in patient self-reporting of symptoms, relapses, and PDDS, NARCOMS provides definitions of terminologies used (i.e. relapse and symptom worsening) on each survey. Further, the six month lag between symptom worsening and relapse reported in two semi-annual surveys separately should have minimized the potential for confusion of these terms. We feel that these steps greatly mitigate the limitations of self-reported data.
Certainly some patients were explicitly internally inconsistent in their responses. For example, despite having reported that their MS symptoms had worsened in the previous 6 months, 66 participants responded to this question with “no change”, “a little better” or “better. This inconsistency shows that the intended definitions do not always seem to match how patients use them. However as a sensitivity analysis, we performed all analyses after excluding these 66 patients; we found that the results were unchanged (results not shown).
In our study, patient-reported relapses were not confirmed by a clinician. Confirming relapses via clinician was not done because it is not feasible to verify each relapse of every patient in such a large self-reported MS registry. The Consortium of Multiple Sclerosis Centers established NARCOMS in 1993 and over 38,000 persons with MS have been registered in this registry with 5–10,000 responding to any one survey. That said, we acknowledge that there are some reported relapses that might have been excluded if a clinician were confirming these.
It is also necessary to differentiate between symptom worsening and disability. In our study, symptom worsening will not necessarily correspond to an increase in disability. In other words, it is possible that symptom could worsen but not be associated with a measureable increase in disability. A patient may report symptom worsening but not a decrease in abilities, especially for those patients with high PDDS scores as is true with EDSS scores. Therefore, SW is not synonymous with disability progression.
Because we ask participants to report symptom worsening that was “not due to relapses or exacerbations”, we would expect relapses not to be predictive of future symptom worsening as it is defined in NARCOMS. And we did, in fact, find no evidence that relapses can predict SW. This indicates at least on face validity that patients did understand the definitions of relapse and symptom worsening they were given and answered accordingly. Furthermore, it is feasible that the significant relationship between symptom change and relapse activity is at least partially mediated by DMT status, with those on optimal treatment potentially experiencing decreased levels in both symptom change and subsequent relapses while those with a sub-optimal treatment response or no treatment having higher levels in both. However, treatment status was not statistically significant in our models so it was not retained in subsequent analysis models.
Here, we used the average of the Performance Scales as one of our measures of symptom change. While using the average allows us to get a global picture of each participant's functioning at each survey, the average may not detect all types of symptom changes and may obscure some changes by averaging that are occurring in only a minor manner in one domain. For example, subjects who improve in one domain but worsen on another may not see a corresponding change in APS. Future work would include considering each PS individually rather than using the average. This approach was not taken here because our data covers a relatively short time frame and PS may change more slowly for more severe presentations of MS.
Nonetheless, this study had several strengths, including its large size, prospective collection of data over five years, contemporary collection of data (all in the last nine years), and use of patient-centered outcomes.
These findings suggest that changes in MS patients' symptoms may precede and predict clinical relapse and future disability progression. Thus, systematic monitoring of patient-reported symptoms may play a useful role in disease management and possibly even treatment decisions. If confirmed by additional studies, a more in-depth evaluation of the potential role of symptom change in predicting future disease activity and disability progression should be pursued.
NARCOMS is supported in part by the Consortium of Multiple Sclerosis Centers (CMSC) and its Foundation. This analysis was supported by Novartis Pharmaceuticals Corporation.
Yuliang Liu participated in the design of the study, conducted statistical analysis and drafted the manuscript.
Charity Morgan participated in the design of the study, conducted statistical analysis and drafted the manuscript.
Lindsey Hornung participated in the design of the study, conducted statistical analysis and revised the manuscript.
Tuula Tyry participated in the design of the study, drafted and revised the manuscript.
Amber Salter participated in the design of the study and revised the manuscript.
Neetu Agashivala participated in the design of the study and revised the manuscript.
Daniel Belletti participated in the design of the study and revised the manuscript.
Edward Kim participated in the design of the study and revised the manuscript.
Robert Fox participated in the design of the study and revised the manuscript.
Stacey Cofield participated in the design of the study, conducted statistical analysis, drafted and revised the manuscript.
Gary Cutter participated in the design of the study, conducted statistical analysis, drafted and revised the manuscript.
Yuliang Liu reports no disclosures.
Charity Morgan reports no disclosures.
Lindsey Hornung reports no disclosures.
Dr. Tuula Tyry reports no disclosures.
Amber Salter has received personal compensation from GlaxoSmithKline for DSMB services.
Neetu Agashivala is employed by Novartis Pharmaceuticals Corporation.
Daniel A Belletti was employed by Novartis Pharmaceuticals Corporation when this study was conducted.
Edward Kim was employed by Novartis Pharmaceuticals Corporation when this study was conducted.
Dr. Fox has received consulting fees from Allozyne, Avanir, Biogen Idec, Novartis, Questcor, and Teva, and research grant support from the National Institutes of Health, National MS Society, and Novartis.
Dr. Cofield serves on data and safety monitoring boards for MedImmune, Orthotech Biotech and consulting fees from the American Shoulder and Elbow Society.
Gary R. Cutter: Dr. Cutter serves on data safety monitoring committees for Antisense Therapeutics Limited, Sanofi-Aventis, Bayhill Pharmaceuticals, Bayer Pharmaceuticals, BioMS Pharmaceuticals, Daichi-Sankyo, Glaxo Smith Klein Pharmaceuticals, Genmab Biopharmaceuticals, Medivation, Peptimmune, PTC Therapeutics, Teva, Vivus, NHLBI, NINDS, National MS Society; and receives consulting or speaking fees from Alexion, Accentia, Barofold, CibaVision, Biogen-Idec, Novartis, Consortium of MS Centers, Klein-Buendel Incorporated, Enzo Pharmaceuticals, Somnus Pharmaceuticals, Teva, Biogen Idec, Advanced Health Media, EMD Serono, EDJ Associates, Aegis Creative Marketing, Eli Lilly, UT Southwestern University, Klein Buendel, University of Illinois Health Policy Center, Somnus Therapeutics.
-  A. Rovira, A. Leon. MR in the diagnosis and monitoring of multiple sclerosis: an overview. Eur. J. Radiol.. 2008;67:409-414 Crossref
-  A.M. Courtney, K. Treadaway, G. Remington, E. Frohman. Multiple sclerosis. Med. Clin. N. Am.. 2009;93:451-476 (ix–x) Crossref
-  Y. Ge. Seeing is believing: in vivo evolution of multiple sclerosis pathology with magnetic resonance. Topics in Magnetic Resonance Imaging: TMRI.. 2006;17:295-306 Crossref
-  L.H. Kasper, J. Shoemaker. Multiple sclerosis immunology: the healthy immune system vs the MS immune system. Neurology. 2010;74(Suppl. 1):S2-S8 Crossref
-  R.J. Fox, F. Bethoux, M.D. Goldman, J.A. Cohen. Multiple sclerosis: advances in understanding, diagnosing, and treating the underlying disease. Cleve. Clin. J. Med.. 2006;73:91-102 Crossref
-  Epidemiology of MS: National MS Society. 2012.
-  E.M. Frohman, M.K. Racke, C.S. Raine. Multiple sclerosis—the plaque and its pathogenesis. N. Engl. J. Med.. 2006;354:942-955 Crossref
-  T.J. Murray. Diagnosis and treatment of multiple sclerosis. BMJ. 2006;332:525-527 Crossref
-  F.D. Lublin, S.C. Reingold. Defining the clinical course of multiple sclerosis: results of an international survey. National Multiple Sclerosis Society (USA) Advisory Committee on Clinical Trials of New Agents in Multiple Sclerosis. Neurology. 1996;46:907-911 Crossref
-  J.H. Noseworthy, C. Lucchinetti, M. Rodriguez, B.G. Weinshenker. Multiple sclerosis. N. Engl. J. Med.. 2000;343:938-952 Crossref
-  R.A. Rudick, E. Fisher, J.C. Lee, J. Simon, L. Jacobs. Use of the brain parenchymal fraction to measure whole brain atrophy in relapsing–remitting MS. Multiple Sclerosis Collaborative Research Group. Neurology. 1999;53:1698-1704
-  J.S. Burks, B.G. Arnason, P.K. Coyle, C.C. Ford, A. Noronha, K.W. Rammohan. Issues and practices in multiple sclerosis. Neurorehabil. Neural Repair. 2002;16:307-320 Crossref
-  R.J. Fox, J.A. Cohen. Multiple sclerosis: the importance of early recognition and treatment. Cleve. Clin. J. Med.. 2001;68:157-171 Crossref
-  E.M. Frohman. Multiple sclerosis. The Medical Clinics of North America. 2003;87:867-897 viii–ix Crossref
-  O.H. Kantarci, B.G. Weinshenker. Natural history of multiple sclerosis. Neurol. Clin.. 2005;23:17-38 v Crossref
-  G. Comi. Shifting the paradigm toward earlier treatment of multiple sclerosis with interferon beta. Clin. Ther.. 2009;31:1142-1157 Crossref
-  B.W. Thrower. Clinically isolated syndromes: predicting and delaying multiple sclerosis. Neurology. 2007;68:S12-S15 Crossref
-  J.A. Cohen, R.A. Rudick. Multiple sclerosis therapeutics. J.A. Cohen, R.A. Rudick (Eds.) Aspects of Multiple Sclerosis That Relate to Trial Design and Clinical Management (Informa Healthcare, United Kingdom, 2007) (p.xxxi, 896 pp.)
-  S. Dhib-Jalbut. Pathogenesis of myelin/oligodendrocyte damage in multiple sclerosis. Neurology. 2007;68:S13-S21 (discussion S43-54) Crossref
-  Secondary Progressive Efficacy Clinical Trial of Recombinant Interferon-Beta-1a in MSSG. Randomized controlled trial of interferon-beta-1a in secondary progressive MS: Clinical results. Neurology. 2001;56:1496-1504
-  M.J. Hohol, E.J. Oray, H.L. Weiner. Disease steps in multiple sclerosis: a simple approach to evaluate disease progression. Neurology. 1995;45:251-255 Crossref
-  M.J. Hohol, E.J. Orav, H.L. Weiner. Disease steps in multiple sclerosis: a longitudinal study comparing disease steps and EDSS to evaluate disease progression. Mult. Scler.. 1999;5:349-354
-  R.A. Marrie, M. Goldman. Validity of performance scales for disability assessment in multiple sclerosis. Mult. Scler.. 2007;13:1176-1182 Crossref
-  C.E. Schwartz, T. Vollmer, H. Lee. Reliability and validity of two self-report measures of impairment and disability for MS. North American research consortium on multiple sclerosis outcomes study group. Neurology. 1999;52:63-70
-  M.P. Sormani, D.K. Li, P. Bruzzi, et al. Combined MRI lesions and relapses as a surrogate for disability in multiple sclerosis. Neurology. 2011;77:1684-1690 Crossref
-  F.D. Lublin, M. Baier, G. Cutter. Effect of relapses on development of residual deficit in multiple sclerosis. Neurology. 2003;61:1528-1532 Crossref
-  C.J. Morgan, A. Ranjan, I.B. Aban, G.R. Cutter. The magnetic resonance imaging ‘rule of five’: predicting the occurrence of relapse. Mult. Scler.. 2013;19:1760-1764 Crossref
-  Interferon beta-1b in the treatment of multiple sclerosis: final outcome of the randomized controlled trial . The IFNB Multiple Sclerosis Study Group and The University of British Columbia MS/MRI Analysis Group (Ed.) Neurology. 1995;45:1277-1285
-  R.A. Marrie, G. Cutter, T. Tyry. Substantial adverse association with visual and vascular comorbidities on visual disability in multiple sclerosis. Mult. Scler.. 2011;17:1464-1471
-  A.L. Cheville, C. Wang, P. Ni, A.M. Jette, J.R. Basford. Age, sex, and symptom intensity influence test taking parameters on functional patient-reported outcomes. Am. J. Phys. Med. Rehabil.. 2014;93:931-937 Crossref
APS - average of Performance Scales
DMT - disease modifying therapy
DP - disability progression
MS - multiple sclerosis
PDDS - Patient-Determined Disease Steps
PRO - patient reported outcome
RA - relapse activity
SW - symptom worsening
a Department of Biostatistics, University of Alabama at Birmingham, Birmingham, AL, USA
b Division of Neurology, Barrow Neurological Institute, Phoenix, AZ, USA
c Novartis Pharmaceuticals Corporation, East Hanover, NJ, USA
d Mellen Center for MS, Cleveland Clinic, Cleveland, USA
⁎ Corresponding author at: University of Alabama at Birmingham, School of Public Health, Department of Biostatistics, 1665 University Blvd. Room 410b, Birmingham, AL 35294-0022, USA.
© 2016 Elsevier B.V., All rights reserved.