Abstract
Background: Primary progressive multiple sclerosis (PPMS) carries the worst prognosis of the multiple sclerosis (MS) subtypes and is currently untreatable. A previous analysis of the British Columbia MS database challenged the view that disability progression is rapid in PPMS, but identified few predictors of disease progression. Here, we extend previous analyses in an updated PPMS retrospective cohort study of prevalent cases.
Methods: We used Kaplan-Meier survival analyses and Cox regression models to investigate the influence of gender, age at onset, and onset symptoms on time to and age at Expanded Disability Status Scale (EDSS) 6.0 in patients with PPMS.
Results: Of 5,779 patients with definite MS, 552 (10%) had PPMS. Median time to EDSS 6.0 was 14.0 years (95% confidence interval [CI] 11.3–16.7), reached at a median age of 58.6 years (95% CI 56.8–60.3). Sensory onset symptoms were associated with a longer time to and an older age at EDSS 6.0 (multivariable hazard ratios 0.55 [95% CI 0.35–0.87] and 0.54 [0.35–0.85]). Younger age at disease onset was associated with a longer time to but a younger age at EDSS 6.0. Gender and other onset symptoms were not associated with these outcomes. Fifty patients with PPMS (9%) fulfilled criteria for benign MS (EDSS ≤3.0 after 10 years' disease duration).
Conclusions: We identified 2 predictors of a slower disease progression in primary progressive multiple sclerosis. Sensory onset symptoms were associated with both a longer time to and a higher age at Expanded Disability Status Scale (EDSS) 6.0. A younger age at disease onset was associated with a longer time to EDSS 6.0, but patients with an early disease onset reached EDSS 6.0 at a younger age.
Glossary
- BCMS=
- British Columbia Multiple Sclerosis;
- CI=
- confidence interval;
- EDSS=
- Expanded Disability Status Scale;
- MS=
- multiple sclerosis;
- PPMS=
- primary progressive multiple sclerosis.
Primary progressive multiple sclerosis (PPMS) is the disease course with the worst prognosis. Although immunomodulatory drugs offer some benefit in relapsing-remitting multiple sclerosis (MS), there is currently no effective treatment for PPMS.1
The long-term natural history of PPMS has been studied previously,2-6 but the relative rarity of this disease course often limits studies to small cohorts, which makes the investigation of possible predictive factors more difficult than in relapsing-remitting MS.
There is currently some uncertainty about the speed of disability accumulation in PPMS: whereas earlier studies showed a rapid disease progression, a more recent analysis of the British Columbia MS cohort showed that the course of PPMS is much slower than previously reported.5 This is important because the estimates of disease progression from natural history studies are used in the design of randomized controlled trials. A recent large randomized controlled trial failed in part because the investigators relied on earlier natural history studies and expected a quicker disease progression than actually took place.7
Many natural history studies have used the time to specific disability levels as the measure of general prognosis, but because MS is a lifelong disease with a course that often spans several decades, it is also important to consider the age at which patients reach certain disability levels.
For this study, we selected a large sample of patients with PPMS to investigate possible predictive factors influencing disability accumulation and age at disability levels in PPMS.
METHODS
Clinical information.
The clinical information was taken from the British Columbia MS (BCMS) database, a longitudinal database linking the 4 MS clinics in British Columbia and capturing more than 80% of the MS population of the province of British Columbia, Canada.8,9 For this retrospective cohort study of prevalent cases, we selected all patients with a diagnosis of clinically definite MS according to the Poser diagnostic criteria10 who were enrolled in the database from September 1980 to July 31, 2003, and had a recorded disability level lower than Expanded Disability Status Scale (EDSS) 6.0 at their first BCMS clinic visit. PPMS disease course was defined clinically as a progressive course from onset without distinct relapses throughout the course of the disease.11 Gender, age at disease onset, and onset symptoms (motor, sensory, optic neuropathy, and cerebellar/ataxia/brainstem), recorded at the patient's first BCMS clinic visit, were also extracted from this database. Age at onset was divided into 4 categories: <30 years, 30 to <40 years, 40 to <50 years, and ≥50 years.
The main outcome measure was time from disease onset to reaching a sustained disability level of 6.0 on the EDSS12 or Disability Status Scale (used until the mid-1980s)13; both scales were used interchangeably. The majority of disability scores were recorded prospectively on standardized forms as part of routine clinical care at each clinic visit by the treating neurologist; a minority were determined retrospectively from patient records. The disability level of EDSS 6.0 was ruled “sustained” if each score following the initial 6.0 score was equal or higher, and if at least 1 follow-up score was taken at least 150 days after the initial 6.0 score.
As a further analysis, we determined the number of patients who fulfilled the criteria for “benign MS” (an EDSS score of 3.0 or below with at least 10 years' disease duration).
Statistical analysis.
The time from disease onset to sustained EDSS 6.0 and the age at sustained EDSS 6.0 were estimated from Kaplan-Meier survival analyses. Group differences in survival time were compared with log-rank tests. In addition, we built multivariable Cox regression models with the time to EDSS 6.0 and age at EDSS 6.0 as the dependent variable. Covariates included in these models were gender, onset symptoms, and age at disease onset for time to EDSS 6.0, and gender and onset symptoms for the outcome age at EDSS 6.0. The age at disease onset was not included as a covariate in the model when estimating the age at EDSS 6.0 because its value is necessarily smaller than the outcome in every case. The model was stratified by age at disease onset instead. We tested for possible violations of the proportional hazards assumption in these multivariable models with log-minus-log plots and Schoenfeld residuals.
We performed a sensitivity analysis with a “most conservative” and “least conservative” scenario, whereby patients excluded from the cohort because they had already reached EDSS 6.0 were included in the analyses with their time to sustained EDSS entered as the time of disease onset (most conservative scenario) or the time of their last follow-up visit (least conservative scenario).
Although this study was not designed to address of a birth cohort effect, i.e., the possibility that the disease course may be different for patients born at different times, we performed an exploratory analysis by categorizing patients as born before 1930, born 1930 to 1939, born 1940 to 1949, and born 1950 and later, and compared the time to EDSS 6.0 between these groups.
All statistical analyses were performed with the statistical software package SPSS version 16 (SPSS Inc., Chicago, IL).
Standard protocol approvals, registrations, and patient consents.
The study was approved by the Clinical Research Ethics Board of the University of British Columbia. All patients gave their written informed consent for the use of their data.
RESULTS
Patient cohort.
Of 5,779 patients with definite MS, 552 (10%) had PPMS. Of these, 109 patients were ineligible for this cohort study because they had already reached a disability level of 6.0 before their first visit. Nineteen of the eligible patients were excluded from the main analyses of time to EDSS 6.0 because they had too few recorded EDSS scores to determine the endpoints. The remaining 424 patients were included in the main analyses. The demographics of included and excluded patients are shown in table 1. There were more women among the excluded patients, but there were no differences in the duration of follow-up, the time to first clinic visit, and age at onset between included and excluded patients (table 1).
Table 1 Comparison of demographic factors between the study cohort and excluded patients
Only 8 patients were exposed to an immunomodulatory treatment before reaching EDSS 6.0 or last follow-up (censoring): 5 were treated with glatiramer acetate, 2 were treated with mitoxantrone, and 1 was treated with cladribine. Given the small number of patients exposed to immunomodulatory treatment, we did not explore the influence of these treatments in this study. Four patients had unknown onset symptoms and were included in the analyses with each of the onset symptoms marked as absent; excluding the patients with unknown onset symptoms did not change the results. A retrospective time to EDSS 6.0 was assigned to 113 patients (27%).
Kaplan-Meier survival analyses.
The results of the Kaplan-Meier survival analyses are given in table 2. The median time to EDSS 6.0 was 14.0 years (95% confidence interval [CI] 11.3–16.7), and the median age at EDSS 6.0 was 58.6 years (95% CI 56.8–60.3). Patients with a younger age at onset had a longer time to EDSS 6.0 (p = 0.004; figure 1). Patients with sensory onset symptoms (p = 0.008; figure 2) and patients without motor onset symptoms (p = 0.02) both took longer to reach EDSS 6.0. The sensitivity analysis showed no substantial difference for these analyses between the observed, most conservative and least conservative scenarios for time to and age at sustained EDSS 6.0 (table e-1 on the Neurology® Web site at www.neurology.org). The comparison of time to EDSS 6.0 by categorized date of birth showed no evidence of a birth cohort effect (table e-2).
Table 2 Kaplan-Meier estimates of time to and age at EDSS 6.0 by potential risk factors in primary progressive multiple sclerosis
Figure 1 Kaplan-Meier curves showing time to and age at sustained EDSS 6.0 by age at disease onset
Time to (left) and age at (right) sustained Expanded Disability Status Scale (EDSS) 6.0.
Figure 2 Kaplan-Meier curves showing time to and age at sustained EDSS 6.0 by sensory onset symptoms
Time to (left) and age at (right) sustained Expanded Disability Status Scale (EDSS) 6.0.
A younger age at onset was associated with a younger age at EDSS 6.0 (p < 0.0005; figure 1). Patients with sensory onset symptoms were older at sustained EDSS 6.0 (p = 0.02; figure 2).
Multivariable regression models.
The results of the multivariable Cox regression models are shown in table 3. The presence of sensory onset symptoms was associated with a longer time to and an older age at EDSS 6.0 (multivariable hazard ratios 0.55 [95% CI 0.35–0.87] and 0.54 [0.35–0.85]). Younger age at disease onset was associated with a longer time to EDSS 6.0 (0.55 (0.35–0.87). Gender and other onset symptoms were not associated with these outcomes. Our investigation of all models with log-minus-log plots and Schoenfeld residuals showed no evidence of a violation of the proportional hazards assumption.
Table 3 Multivariable analyses of factors affecting time to and age at EDSS 6.0
Variability in disease progression.
There was a marked variability in survival estimates. Whereas 25% of patients took less than 7.8 years to reach EDSS 6.0 and at an age of 49 years (first quartile estimated from the Kaplan-Meier curves), the 25% with the slowest progression reached this outcome after more than 27 years and at an age of 70.2 years (last quartile).
“Benign PPMS.”
Most patients (381/533, 69%) had reached or passed the level of EDSS 3.0 before first seen at the MS clinic. We identified 50 patients (9% of the whole cohort of 552 patients) who fulfilled the criteria for “benign MS.”
DISCUSSION
In this study, we found considerably slower progression to EDSS 6.0 than reported in previous cohorts. In contrast to the 7.1 years in the Lyon cohort (n = 282)3 or approximately 8.5 years in the London, Ontario cohort (n = 216),4 our MS population took 14 years to reach EDSS 6.0. This considerable difference between cohorts is difficult to understand. One explanation could be different environmental effects that result in one local form of PPMS being more aggressive than the other. Alternatively, because our cohort is more recent than the other 2 large population-based cohorts, the difference could reflect a temporal change in the progression of PPMS. Interestingly, a recent trial in PPMS7 found progression rates similar to those in our previous investigation,5 which provides support for the suggestion that the course of PPMS may be changing; conversely, our exploratory analysis on a possible birth cohort effect showed no evidence for such an effect. This merits further investigation. Other potential explanations for the difference between our results and those from other cohorts include a referral bias in the other cohorts, which may have resulted in the inclusion of more severe cases, and differences in the definition of PPMS; our PPMS cohort includes only patients without distinct relapses, whereas previous studies have also included patients with relapsing-progressive disease. Previous studies have, however, shown similar disease progression rates in PPMS and progressive-relapsing MS.3,14
Few predictors of disease progression in PPMS have been found. In our cohort, we were able to report on 2 factors affecting disease outcome. Contrary to what has been reported previously, we found that sensory onset symptoms were associated with a better prognosis. In a previous analysis of the BCMS cohort, we found a similar difference in median time to EDSS 6.0 for sensory onset symptoms, but this difference was not significant, probably because of a smaller sample size.5 The absence of motor onset symptoms was associated with a longer time to EDSS 6.0, but this association did not remain significant in the multivariable models. We also found that a younger age at disease onset was associated with a longer time to EDSS 6.0. However, our investigation of the age at EDSS 6.0 showed that patients who were younger at disease onset also required a cane at a younger age. A young age at disease onset should therefore not be considered predictive of a better prognosis.
Although PPMS carries the worst prognosis of all MS subtypes, it is important to realize that there is a large variation in the time to EDSS 6.0 among patients with PPMS. In our cohort, 25% of patients with the quickest disease progression reached EDSS 6.0 after less than 8 years (the first quartile estimated from the Kaplan-Meier curve), whereas the 25% with the slowest disease progression took more than 27 years to reach this level of disability (last quartile). Thus, although the overwhelming majority of patients with PPMS became moderately disabled eventually, the time from disease onset until patients require the use of a cane can be substantial. This information can be useful for the counseling of patients with PPMS.
The large number of patients in our cohort also enabled us to investigate “benign PPMS.” Interestingly, approximately one-tenth of our cohort fulfilled the criteria for “benign MS” (an EDSS score of 3.0 or less at 10 years' disease duration). Although this is another finding supportive of the idea that the disease course of PPMS is not as aggressive as previously thought, it should be kept in mind that the EDSS evaluation does not capture all physical limitations (such as disabling tremor) or any cognitive impairment or depression.
Limitations of our study include the fact that few patients in our study cohort had optic neuritis or cerebellar and brainstem symptoms at onset, which makes the survival estimates for these possible influencing factors less reliable. Certain patients were excluded from these analyses because of a lack of information or because they were ineligible for this cohort study due to advanced disability at the first clinic visit. The exclusion of these cases seems unlikely to have biased the main analysis, however, because these patients were demographically similar to the included patients, apart from a slight predominance of women (gender was not a predictor of disease progression). Furthermore, the sensitivity analyses with most and least conservative scenarios did not change our estimates substantially or our interpretation. Our geographically and population-based cohort is estimated to contain 80% of the patients with MS in the province of British Columbia; however, these estimates date from the 1980s, and we cannot confidently estimate our current population coverage. It is possible that the 20 or more percent of patients who never attend clinic are either especially mild or especially severe cases, which may influence the reliability of our reported rates of disease progression in PPMS. Another limitation of our study and other studies on PPMS is the difficulty of making a confident diagnosis of PPMS, which often results in a long delay between symptom onset and first clinic visit (more than 7 years on average in our cohort). This delay may impact the patients' ability to recall onset symptoms.
Natural history studies are important to guide the design of randomized controlled trials. Recently, a large trial on the effect of the immunomodulatory treatment glatiramer acetate on disease progression was unsuccessful, in part because a more rapid disease progression was anticipated in the placebo arm.7 The results of this study and the finding that disability accumulation is a much slower process than previously reported will hopefully help in the design of future trials involving patients with this currently untreatable form of MS.
AUTHOR CONTRIBUTIONS
The statistical analyses were performed by Marcus Koch.
ACKNOWLEDGMENT
The statistical advice of Dr. Yinshan Zhao is gratefully acknowledged. The University of BCMS clinic neurologists contributed to this study through data collection. Their contribution is gratefully acknowledged; contributing neurologists were (in alphabetical order) D. Adams, D. Craig, L. Daly, V. Devonshire, S. Hashimoto, O. Hrebicek, J. Hooge, B. Jones, L. Kastrukoff, S. Meckling, J. Oger, D. Parton, D. Paty, P. Smyth, W. Shtybel, and T. Traboulsee.
DISCLOSURE
Dr. Koch receives research support from the Canadian Institutes of Health Research (CIHR) and from the Multiple Sclerosis (MS) International Federation. Dr. Kingwell receives research support from the MS Society of Canada. Dr. Rieckmann serves as Research Chair of the MS Society of Canada; serves on scientific advisory boards for the Germany MS Society, Merck Serono, Novartis, Teva Pharmaceutical Industries Ltd., and Bayer-Schering Pharma; serves on the editorial advisory board of Therapeutics in Neurology; has received speaker and/or consulting honoraria from Bayer-Schering Pharma, Biogen Idec, Merck Serono, Novartis, and Teva Pharmaceutical Industries Ltd.; and receives research support from the MS Society of Canada, the US National MS Society, and the CIHR. Dr. Tremlett has received speaker honoraria from the Swiss MS Society and the University of British Columbia MS Research Program; and has received research support from the CIHR [190898 (PI) and MOP-82738 (PI)], the US National MS Society, and the MS Society of Canada (Don Paty Career Development Award), and is a Michael Smith Foundation for Health Research Scholar. The BCMS database was funded by an unrestricted grant from Dr. Donald Paty and the MS/MRI Research Group.
Footnotes
-
Supplemental data at www.neurology.org.
Disclosure: Author disclosures are provided at the end of the article.
Received June 5, 2009. Accepted in final form September 15, 2009.
REFERENCES
Letters: Rapid online correspondence
REQUIREMENTS
You must ensure that your Disclosures have been updated within the previous six months. Please go to our Submission Site to add or update your Disclosure information.
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.
