Abstract
Background: Ten percent of patients with MS have a progressive course from onset with no history of relapses or remissions. A smaller subgroup follow a similar progressive course but have a single relapse at some point (transitional progressive [TP] MS). To date these patients have been excluded from receiving licensed treatments for MS and from most therapeutic trials.
Objective: To document the clinical and MRI characteristics of a large cohort of progressive patients, including 158 with primary progressive (PP) MS and 33 with TPMS. Data from a small reference group of 20 patients with secondary progressive (SP) MS are also presented for reference.
Methods: Patients were recruited from six European centers. All underwent a clinical assessment including scoring on the Expanded Disability Status Scale (EDSS) and MRI of the brain and spinal cord.
Results: The men-to-women ratio was 81:77 (51% men) in the PP group, 14:19 (42% men) in the TP group, and 5:15 (25% men) in the SP group. The mean age at disease onset was significantly higher in the PP group than it was in the other two groups (PP 40.2 years, TP 34.9 years, SP 28.7 years). On MRI the PP group had lower mean brain T2 and T1 hypointensity lesion loads than the SP group (T2 12.02 versus 27.74 cm3, p = 0.001; T1 4.34 versus 7.04 cm3, p = 0.015). The SP and TP cohorts had significantly more T2-weighted lesions in the spinal cord than the PP patients, and the SP cohort had the greatest degree of atrophy. There was a correlation in the PP and TP patients between EDSS score and brain and spinal cord atrophy (r = 0.3, 0.2, p ≤ 0.006) but not with brain lesion load. The PP and TP patients who presented with spinal cord pathology had significantly lower brain T2 and T1 lesion loads than those with non-spinal cord presentations (p = 0.002).
Conclusions: The monitoring of disease progression in PPMS is difficult, although measures of atrophy correlate with the EDSS and appear most promising. This study increases our understanding of this unique patient group, which will be further expanded with the acquisition of serial data.
The majority of patients with MS present with a relapsing-remitting (RR) course. However, many of these patients subsequently enter a progressive phase with or without superimposed relapses—secondary progressive (SP) MS. Disability accumulates as a consequence of both incomplete remission after a relapse and gradual disease progression. Recent therapeutic trials have concentrated on these patient groups and have looked at outcome measures that monitor both relapse rate and accumulated disability.1-5 However, 10% of patients with MS have a progressive course from onset with no history of relapse or remission—so-called primary progressive (PP) MS. There are also a few patients with an essentially progressive course with the exception of a single relapse or remission at some time before or after the onset of disease progression. These patients have been labeled as having “transitional progressive” (TP) MS.6-8 The classification of patients with progressive forms of MS is extremely difficult, as highlighted in a recent survey of members of the international MS clinical research community.9 It also included a progressive-relapsing category that is similar to TP, with the important exception that more frequent relapses are allowed. These patients will not be considered here.
Difficulties in the diagnosis of PPMS have recently been highlighted.7,10 The Poser criteria require evidence of disease dissemination in both time and space, although they allow evidence from paraclinical tests (oligoclonal bands on CSF examination, delay in visual evoked potentials, and abnormalities on MRI), which may need to be repeated over time to fulfill the criteria. Unlike patients with RRMS who commonly present with a visual or sensory disturbance, patients with PPMS usually present with a progressive but monosymptomatic course, usually a spastic paraparesis. However, a small proportion will present with progressive cerebellar, brainstem, visual, or hemiplegic syndromes; very rarely progressive cognitive decline can occur.
Few studies have looked at this patient group, and because of their relative rarity the number of patients involved is often small. However, differences between PPMS and the RR and SP subtypes are well documented. The age at onset appears to be later,11,12 and the male-to-female ratio is equal in contrast to the female preponderance seen in RRMS.13,14 Differences in the genetic and immunologic profiles of PPMS have been noted, but the study numbers are small and results inconsistent.7 In the only study to date comparing the cognitive function in patients with PPMS with those with SPMS, cognitive dysfunction was reported in only 7% of patients with PPMS compared with 53% of SPMS patients.15 There has been only one paper published specifically addressing the pathologic findings in PPMS, which detected less inflammatory changes than was seen in the SP group.16 On MRI, brain lesions are seen less frequently in the PP patient group, and those that are detected tend to be smaller.17 When compared with SPMS, fewer new lesions develop per year (3.3 per patient per year compared with 18.2 in the SP group), and of these only 5% enhanced with gadolinium-DTPA compared with 87% in the SP group.18 The MRI findings in TPMS appear similar to those in PPMS.6,8 MRI of the spinal cord has not shown any differences between the MS subtypes in the number or size of lesions,19-21 although diffuse signal change has been noted in patients with PPMS21 and the degree of spinal cord atrophy appears to correlate strongly with disability, particularly in the progressive subgroups.22
At present treatments for MS (interferon beta-1a and 1b and glatiramer acetate) are only licensed for RRMS. Studies addressing the role of treatments in SPMS are nearing completion. To date we are aware of only two small treatment trials for patients with PPMS, one with interferon-beta 1a23 and the other with riluzole (C.H. Polman, personal communication, 1998).
Further evaluation of these relatively rare groups is important for several reasons. Firstly, documenting their clinical course and MRI features will advance our understanding of the evolution of disease in MS and, in particular, will help in elucidating whether the different subgroups, defined on the clinical pattern of disease, are merely parts of the spectrum of MS or are, in fact, distinct disease entities with unique MRI characteristics. Secondly, because disease activity in PPMS is by definition purely progressive, more can be learned about the underlying mechanisms of disability resulting from disease progression as distinct from the effect of relapses. Thirdly, and perhaps most importantly, the documentation of their clinical and MRI characteristics will facilitate appropriate trial design, which at present is hindered by the choice of outcome measures. Most clinical trials in MS rely on the rate of new brain lesions or gadolinium enhancement as a measure of short-term disease activity and on T2 lesion load for chronic phase monitoring.24 The limited studies to date suggest that few new brain lesions are seen, and lesion loads tend to be small. No serial studies assessing change in lesion loads have been reported.
There is thus a need to carry out a more extensive study of this important patient group to address these questions involving the serial evaluation of a large number of patients. This by necessity is of multicenter design. This paper describes the initial cross-sectional results of such a study funded by the EC initiative MAGNIMS (Magnetic Resonance Network in Multiple Sclerosis).25 Neuropsychological results will be reported separately.
Methods.
Patients.
Patients were recruited from six centers (Amsterdam, Barcelona, Bordeaux, Lisbon, London, Milan) with no restriction on age or disability. A careful history was obtained to exclude patients with a nonprogressive disorder or, in relation to TPMS patients, those with more than one relapse in the course of their disease. Patients were classified before entering the study as having either PP or TPMS based solely on their clinical history. Previous MR imaging was not assessed and did not influence categorization of disease type. An additional 20 patients with SPMS were recruited in London, not as a direct comparison but as a reference group to aid comparisons with previously published data. Each patient underwent a clinical and neuropsychological assessment using measures that were familiar to all the participating sites. Impairment and disability were assessed using the Kurtzke Expanded Disability Status Scale (EDSS), a 10-meter timed walk (fastest time in seconds to walk 10 meters with usual aids), and the 9-hole peg test.26
Imaging protocol.
All London and Lisbon scans were carried out using a Signa 1.5-T system (General Electric, Milwaukee, WI). The other four sites used Siemens Magnetom 1.5-T systems (Munich, Germany). Each patient underwent T1- and T2-weighted spin-echo imaging of the brain. All sequences were acquired as contiguous, 3-mm–thick, axial slices (44 images in total). In the spinal cord, nine contiguous, 3-mm, sagittal T2 and proton-density–weighted slices were obtained. A volume-acquired inversion-prepared gradient-echo acquisition of the cervical cord (60 1-mm slices) was also performed (Signa scanners, fast-spoiled gradient echo; Magnetom scanners, magnetization-prepared rapid-acquisition gradient echo), and from the data set a series of five contiguous 3-mm axial slices (perpendicular to the spinal cord) were reformatted using the center of the C2-3 disk as the caudal landmark. The imaging parameters for each site were similar. Details are on file with the National Auxiliary Publications Service (NAPS) (see Note at end of text).
Analysis.
The 9-hole peg test was used as a measure of disability by adding the times from both hands. When the patient was unable to perform the task or took longer than 5 minutes to complete it, the time for that hand was recorded as 300 seconds.
Brain MRI lesions were identified and marked on the proton-density–weighted films with cross reference to the T2-weighted images by one observer who was blinded to patient type and details (D.H.M.). T1 hypointensities or “black holes” (areas of decreased signal intensity demarcated from surrounding tissue and corresponding to identified lesions on the T2-weighted films) were marked in a similar way. T2 and T1 lesion volumes of the brain were then assessed with a semiautomated contouring technique27 by a single observer (V.L.S.).
A measure of partial brain volume reflecting atrophy was acquired as described by Losseff et al.28 This technique measures the volume of brain covered by six contiguous 3-mm slices, with the most caudal one at the level of the velum interpositum cerebri. This site was chosen because it covers a large proportion of the lateral ventricles and cortical sulci, and the velum interpositum cerebri is thought to be a stable landmark despite ongoing atrophy, allowing repositioning for serial assessment.
A single observer (D.H.M.) also marked spinal cord lesions. Unfortunately, at the present time resolution of spinal cord imaging is not sufficient to allow automated measures of lesion load. Instead, the size of each lesion was denoted by multiplying the length of the lesion (the number of vertebral bodies over which the lesion extended) by the number of slices on which the lesion was visible. Total spinal cord lesion load was determined by the sum of the individual lesion sizes.
Spinal cord cross-sectional area at the C2-3 level was measured using a semiautomated technique.22 This level was chosen because there is little variability in cross-sectional area over this segment and the CSF pool is capacious, thus optimizing spinal cord/CSF contrast. Data acquired with the Signa system (London and Lisbon) were analyzed by V.L.S., and data acquired with the Siemens system were analyzed by M.R.
Reproducibility.
Brain lesion load, cerebral atrophy measures, and cross-sectional spinal cord areas were repeated on 10 random subjects 2 weeks apart. The coefficient of variation (COV) was calculated for each measure by dividing the standard deviation by the mean to assess measurement reproducibility.
Statistical analysis.
Nonparametric statistical tests were used throughout. The Mann-Whitney test was used to look for differences between the patient groups. Correlations were assessed using Spearman’s rank correlation coefficient. To reflect the large number of statistical comparisons, a p value of 0.01 was considered significant, and a value between 0.05 and 0.01 was considered a trend. No mathematical correction of statistical significance was carried out to avoid inflating type II errors (the probability of accepting the null hypothesis when the alternative is true) and thus missing real differences.29
Results.
A total of 158 patients with PPMS, 33 with TPMS, and 20 with SPMS were recruited. The men-to-women ratio was 81:77 (51% men) in the PP group, 14:19 (42% men) in the TP group, and 5:15 (25% men) in the SP group. Other clinical characteristics are listed in table 1. The PP patients were significantly older at the onset of their disease than those in the other two groups, whereas the SP group was the most disabled but also had the longest disease duration.
Clinical and MRI findings
Intrarater reproducibility for the COV for brain lesion load analysis was 2.48%; the more automated measures of brain atrophy and cord cross-sectional area produced COVs of 0.65% and 0.51%, respectively.
The PP group had significantly lower mean T2 and T1 brain lesion loads than the SP group (p = 0.001, p = 0.015); the TP group was intermediate. The T2/T1 mean lesion load ratios were similar in the three patient groups (PP 2.8, TP 2.9, SP 3.9). There was no significant difference in the six-slice measure of brain volume between the three patient groups.
In the spinal cord, the PP group had significantly fewer focal lesions and a smaller spinal cord lesion volume than the TP group (p < 0.001, p = 0.008) and fewer spinal cord lesions than the SP group (p = 0.04). Measurement of spinal cord cross-sectional area revealed that the SP group had a significantly smaller area than either the PP group (p = 0.001) or TP group (p = 0.007).
The entire group showed an extremely weak correlation between brain T2 lesion load and both the EDSS (r = 0.16, p < 0.03) and disease duration (r = 0.16, p < 0.03). However, these relationships were not apparent in any of the individual disease groups. Spinal cord cross-sectional area measurements correlated more strongly with the EDSS (r = −0.30, p < 0.001). The clinical measures of upper-limb function and ambulation both correlated strongly with the EDSS (9-hole peg test: r = 0.59, p < 0.001; 10-meter timed walk: r = 0.66, p < 0.001).
Despite the lack of correlation between lesion volumes and the EDSS, within the PP and TP groups together, the 9-hole peg test did correlate with both T1 (r = 0.33, p < 0.001) and T2 (r = 0.35, p < 0.001) lesion loads. The other significant correlations between MRI and clinical measures were those concerning spinal cord and brain atrophy. Spinal cord cross-sectional area correlated with both the EDSS (r = −0.26, p = 0.001) and the 9-hole peg test (r = −0.33, p < 0.001). The measure of partial brain volume correlated with the 10-meter timed walk (r = −0.39, p < 0.001) and weakly with the EDSS (r = −0.20, p = 0.006). There were no correlations between brain and spinal cord MRI findings or indeed between the number or volume of spinal cord lesions and spinal cord cross-sectional area.
The majority of patients with PPMS and half of the TPMS patients presented with a progressive spinal cord syndrome. The remaining patients and the SP group were more varied (table 2). When the 147 patients presenting with a spinal cord syndrome were compared with the 44 with other presentations, no difference in disease duration or EDSS was seen (table 3). However, the spinal cord group had significantly lower brain T2 and T1 lesion loads (p = 0.002). There was no difference in the spinal cord MRI measures between the two groups, although there was a significant correlation between spinal cord cross-sectional area and the EDSS within the spinal cord group (r = −0.27, p = 0.002) that was not present in the other group (r = −0.19, p = 0.25). As expected, the spinal cord-onset group scored worse on the pyramidal (p = 0.04) and sensory (p = 0.02) functional system scores of the EDSS and the 10-meter timed walk (p = 0.04), but better on the cerebellar (p = 0.001) and brainstem (p = 0.05) functional scales as well as the 9-hole peg test (p = 0.03).
Presentation at disease onset
Presentation at onset—primary and transitional progressive patients
All the SP and two-thirds of the TPMS patients were recruited in London, making an intersite comparison unfeasible; however, this was possible for the PPMS patients. The mean age at disease onset and rate of disease progression (EDSS/disease duration) were similar at each site. The Barcelona cohort had the largest mean T2 brain lesion load, whereas the London cohort had more spinal cord lesions and larger spinal cord lesion volumes than the other centers. The mean spinal cord cross-sectional area measurements revealed significant differences between the Barcelona cohort (the larger value) and the London and Lisbon cohorts (further details available through NAPS; see Note at end of text).
Discussion.
This study confirms that patients with PPMS differ clinically from the other subtypes of progressive MS in several ways. The late age at onset (40.2 years) is in agreement with previous studies11,12,30,31 and is significantly later than all other subgroups. The incidence appears to be equal in men and women. This lack of female preponderance, which is usually seen in autoimmune conditions, may reflect the less inflammatory/immune nature of the disease course compared with other subgroups of MS.
The MRI findings in this study are also consistent with previous findings.17 Patients with PPMS had low brain T2 and T1 lesion loads, and no correlation was found with the EDSS. As with other studies, the number and volume of spinal cord lesions differed little between the subgroups, and there was no correlation with disability.19-21
The prognosis in PPMS is thought to be poor in relation to disability.13,31 This is supported by the findings in this study. Although the progression index (EDSS/disease duration) does not differ significantly between the three patient groups, it is highest in the PP group, with the TP group intermediate to the PP and SP groups. This finding is similar to that of previous investigators8 and occurs despite low numbers of inflammatory lesions in the brain or spinal cord (pathologically16 or on MRI), suggesting that the mechanism of disability in PPMS cannot be solely attributed to lesion formation.
The only MRI measures correlating significantly with the EDSS are those reflecting atrophy of both the brain and spinal cord. Atrophy occurs secondary to a combination of demyelination32 and axonal loss.33 In patients with PPMS the development of disability may be a consequence of diffuse disease of the normal-appearing white matter (NAWM), resulting in atrophy. This hypothesis is supported by recent MR spectroscopy showing reduced levels of an axonal marker, N-acetyl aspartate (NAA), in the NAWM of patients with PPMS.34-36 Fu et al.37 also showed a correlation between change in the EDSS and the NAA:creatine ratio in the NAWM of patients with RRMS. No such correlation was seen between the T2 lesion load and the EDSS.37 This suggests that diffuse axonal loss may be an important factor in the accumulation of disability in MS even in early RRMS. Areas of NAWM have also been studied using magnetization transfer (MT) imaging to assess the degree of demyelination present. The MT ratio was significantly lower in patients with chronic progressive MS than it was in patients with RR disease, although patients with PPMS were not distinguished from SPMS.38,39
Most of the patients with PPMS (83%) and nearly half of the TPMS patients (49%) presented with a progressive spastic paraparesis. These patients did not differ from other presentations in either disease duration or disability as measured by the EDSS. However, their brain T2 and T1 lesion loads were significantly and considerably lower than the others (although no difference in the number or volume of spinal cord lesions was seen). Despite there being no difference between the spinal cord presentation group and the others in spinal cord cross-sectional area, there was a significant correlation with the EDSS within the spinal cord group that was not present in the other group. This may reflect the small sample size in the other presentation group or the influence of ataxia or brainstem scores on the EDSS in this group.
On comparing the PP patients between centers the clinical characteristics are similar, with the main disparity being the MRI results. In the spinal cord this perhaps reflects the different scanners used. London and Lisbon both imaged using a GE Signa scanner, and more lesions were identified. The protocol was designed in London and may therefore be optimized for these two centers. The variation between brain lesion loads may be a result of small sample size in some of the centers, although this would not explain the results of the Barcelona cohort, who had larger brain lesion volumes despite comparable disease durations and level of disability. It would be of interest to assess an SP cohort in each center to see if similar variations occur in this patient group. Variations in measured brain lesion loads between different scanners may also play a part, particularly if the field strengths differ,40 although these effects are small provided patients are followed up on the same scanner. However, if measures of atrophy, where extremely small changes are expected to occur, are to be used in the monitoring of clinical trials, the intrascanner variations may be more important. These include both day-to-day fluctuations of field and gradient strength and, in particular, the effects of servicing or upgrading the machine. Losseff et al.41 observed an increase in the size of the spinal cords of control subjects and some patients over a 12-month period as a consequence of a shift in the gradients after an upgrade, thus emphasizing the need for resilient quality assurance (QA) programs in all units and for all clinical trials. After this episode a QA program was established in London to detect any further fluctuations.42 Interestingly, the correlation between the EDSS and cross-sectional spinal cord measurements of the PP patients in London (where the reliability of spinal cord measurements is ensured by QA) is strong (r = −0.43, p = 0.001), but when all the other centers are combined it is poor and insignificant (r = −0.16, p = 0.18). This may reflect the larger number of patients in this single center or perhaps suggests there are undetected variations in the acquisition of the spinal cord volume data across sites. All centers are now regularly scanning controls to investigate this further.
These findings are important both in understanding the disease process but also in the design of therapeutic trials. At present the disease subgroups of MS are defined by clinical course. Despite the evidence that the PP group show marked differences in their clinicopathologic features, the results of this study demonstrate that there is still considerable overlap in both the brain and spinal cord findings between MS subtypes preventing distinction on radiologic grounds. In fact, when this large cohort is separated into modes of presentation, one subgroup of PPMS patients demonstrates very large brain lesion loads, approaching those of the SPMS group. This may in fact suggest that our clinical classification of MS subtypes does not represent distinct disease entities but merely represents different portions of a spectrum of MS characteristics. This theory is further supported by the findings in patients with TPMS, a group that has many similarities to PPMS patients but with MRI findings midway between PP and SPMS.
Trial design in PPMS is extremely difficult because appropriate outcome measures are as yet unknown. The lack of correlation between the EDSS and lesion loads in this large study is of concern and whether the T2 or T1 lesion loads change significantly over the time span of a clinical trial is yet to be demonstrated. However, the correlations between measures of atrophy and the EDSS, 9-hole peg test, and 10-meter timed walk are encouraging. The rarity of PPMS means definitive trials will inevitably be of multicenter design, this study has also highlighted problems with such studies particularly if measures such as atrophy, where extremely small degrees of change are detected, are to be relied upon. This large and valuable cohort will be followed annually to address these issues further.
Acknowledgments
Acknowledgment
The authors thank H. Brochet, G. Comi, L. Herisse-Dulou, R. Hintzen, P. Lafon, and J. Rio for their contribution in making this study possible.
Note. Readers can obtain 2 pages of supplementary material from the National Auxiliary Publications Service, 248 Hempstead Turnpike, West Hempstead, NY 11552. Request document no. 05491. Remit with your order (not under separate cover), in US funds only, $15.00 for photocopies and $5.00 for microfiche. Outside the United States and Canada, add postage of $4.50 for the first 20 pages and $1.00 for each 10 pages of material thereafter, or $1.75 for the first microfiche and $1.00 for each fiche thereafter. There is a $25.00 invoicing charge on all orders filled before payment.
Footnotes
This work was part of MAGNIMS (Magnetic Resonance Networks in Multiple Sclerosis), funded by an EC initiative. The Institute of Neurology NMR Research Unit and V.L.S.’s post are funded by the Multiple Sclerosis Society of Great Britain and Northern Ireland. The portion of the study conducted in Bordeaux, France was supported by a grant from Ligue Française Contre la Sclerose en Plaques.
- Received July 16, 1998.
- Accepted in final form November 21, 1998.
References
Disputes & Debates: Rapid online correspondence
REQUIREMENTS
If you are uploading a letter concerning an article:
You must have updated your disclosures within six months: http://submit.neurology.org
Your co-authors must send a completed Publishing Agreement Form to Neurology Staff (not necessary for the lead/corresponding author as the form below will suffice) before you upload your comment.
If you are responding to a comment that was written about an article you originally authored:
You (and co-authors) do not need to fill out forms or check disclosures as author forms are still valid
and apply to letter.
Submission specifications:
- Submissions must be < 200 words with < 5 references. Reference 1 must be the article on which you are commenting.
- Submissions should not have more than 5 authors. (Exception: original author replies can include all original authors of the article)
- Submit only on articles published within 6 months of issue date.
- Do not be redundant. Read any comments already posted on the article prior to submission.
- Submitted comments are subject to editing and editor review prior to posting.