Clinical Correlation of Multiple Sclerosis Immunopathologic Subtypes

Abstract

Background and Objectives

The goal of this work was to compare clinical characteristics across immunopathologic subtypes of patients with multiple sclerosis.

Methods

Immunopathologic subtyping was performed on specimens from 547 patients with biopsy- or autopsy-confirmed CNS demyelination.

Results

The frequency of immunopathologic subtypes was 23% for pattern I, 56% for pattern II, and 22% for pattern III. Immunopatterns were similar in terms of age at autopsy/biopsy (median age 41 years, range 4–83 years, p = 0.16) and proportion female (54%, p = 0.71). Median follow-up after symptom onset was 2.3 years (range 0–38 years). In addition to being overrepresented among autopsy cases (45% vs 19% in biopsy cohort, p < 0.001), index attack-related disability was higher in pattern III vs II (median Expanded Disability Status Scale score 4 vs 3, p = 0.02). Monophasic clinical course was more common in patients with pattern III than pattern I or II (59% vs 33% vs 32%, p < 0.001). Similarly, patients with pattern III pathology were likely to have progressive disease compared to patients with patterns I or II when followed up for ≥5 years (24% overall, p = 0.49), with no differences in long-term survival, despite a more fulminant attack presentation.

Conclusion

All 3 immunopatterns can be detected in active lesions, although they are found less frequently later into the disease due to the lower number of active lesions. Pattern III is associated with a more fulminant initial attack than either pattern I or II. Biopsied patients appear to have similar long-term outcomes regardless of their immunopatterns. Progressive disease is less associated with the initial immunopattern and suggests convergence into a final common pathway related to the chronically denuded axon.

Multiple sclerosis (MS) exhibits significant heterogeneity in clinical,¹ radiographic,² and pathologic findings.³ Pathologic hallmarks include multifocal demyelination, inflammation, gliosis, and axonal damage. Demyelinating activity can be staged according to the presence of different myelin degradation products within macrophages, including early active, late active, inactive, early remyelinating, and late remyelinated (shadow) plaques.³ Four distinct pathologic forms of the acute MS plaque have been described in the context of an acute demyelinating lesion, with 3 commonly observed (Figure 1). These immunopatterns differ in the effector mechanism of demyelination. Specifically, immunopattern I is associated predominantly with T-cell/macrophage infiltration; in immunopattern II, complement deposition plays a significant role; immunopattern III represents a distal oligodendrogliopathy, with preferential loss of myelin-associated glycoprotein and oligodendrocyte apoptosis; and immunopattern IV is considered a primary oligodendrocyte degeneration.

Figure 1. Composite Slide of Immunopatterns Seen in Active Demyelination.

Tissue immunohistochemistry of myelin proteolipid protein (PLP), myelin-associated glycoprotein (MAG), myelin oligodendrocyte glycoprotein (MOG), and complement products (C9neo), demonstrating typical findings in immunopattern I (A–D), II (E–H), and III (I–L).

C9neo = C9neoantigen; IP = immunopattern; MAG = myelin-associated glycoprotein; MOG = myelin oligodendrocyte glycoprotein; PLP = proteolipoprotein.

Immunopatterns of early active MS plaques are patient specific. They reflect between-patient heterogeneity but within-patient homogeneity across all their actively demyelinating lesions.⁴ The relationship between clinical outcomes and pathologic immunopattern is unclear. Because each patient has 1 specific immunopattern in all active demyelinating plaques throughout the course of the disease, it is hypothesized that these immunopatterns may be correlated with differences in presentation or outcome of MS. Differences in cross-sectional findings and long-term clinical outcomes of patients with MS with respect to the pathologic findings of the acute MS plaque are presented here.

Methods

Standard Protocol Approvals, Registrations, and Patient Consents

Study approval was granted by the Institutional Review boards of Mayo Clinic, Rochester, MN (No. 2067-99), Medical University of Vienna, Austria (EK No. 535/2004/2019), and University Medical Center Göttingen, Germany (No. 19/09/10). Written informed consent was obtained from all participants before they joined the study. We use no figures or videos with recognizable participants.

This study was performed on patients with histopathologically proven active demyelination determined via biopsy or autopsy who were enrolled in the Multiple Sclerosis Lesion Project.^3,5 Material was collected in the Department of Laboratory Medicine and Pathology (Neuropathology) at the Mayo Clinic, Rochester, MN; the Neuropathologic Institute at the University of Göttingen, Germany; and the Institute of Brain Research at the University of Vienna, Austria. The inclusion criteria for cases in this study were as follows: (1) tissue diagnosis of inflammatory demyelination confirmed by a specialist in neuropathology of demyelinating disease (W.B., J.E.P., Y.G., H.L., I.M.,C.F.L.) to be consistent with MS, with the presence of confluent plaques in active stage of myelin destruction and relative sparing of axons and glial scaring; (2) at least 1 lesion in the active stage of demyelination; (3) no clinical, radiologic, serologic, or pathologic evidence of neoplasm, infection, vascular, or nondemyelinating inflammatory etiology; and (4) no structural or immunocytochemical evidence of an inflammatory demyelinating disease induced by known virus infections such as subacute sclerosing panencephalitis or progressive multifocal leukoencephalopathy. Cases of acute disseminated (perivenous) leukoencephalomyelitis, myelin oligodendrocyte glycoprotein (MOG) antibody–associated disease, and neuromyelitis optica were excluded according to histopathologic and serologic criteria.^6,7 Patients were classified into 1 of 4 distinct immunopathologic patterns, as described previously.³ Note that although pattern IV was originally described with the other immunopatterns, no further cases with this pattern have been found. This article therefore focuses on patterns I through III. Some data from patients presented here have been published as part of prior publications using data from the Multiple Sclerosis Lesion Project.^4,8-13

Neuropathologic Techniques and Immunocytochemistry

All cases underwent detailed neuropathologic examination. All tissue blocks were classified with regard to stage of demyelinating activity.¹⁴ Paraffin-embedded 5-μm sections were stained with hematoxylin & eosin, Luxol fast blue (myelin)/periodic acid–Schiff, and Bielschowsky silver impregnation (axons).

Immunohistochemistry

Immunohistochemistry was performed on paraffin-embedded sections using an avidin-biotin or an alkaline phosphatase/anti–alkaline phosphatase technique as described in detail previously.¹⁵ The primary antibodies were omitted as negative controls. In situ hybridization was performed using digoxigenin-labeled riboprobes specific for proteolipoprotein. The source and specificity of the probes, the labeling techniques, and the methods of in situ hybridization have been described in detail previously.¹⁶ To visualize degenerating cells in tissue sections, DNA fragmentation within cell nuclei was determined with the method of in situ tailing.¹⁷ The sections were then processed for immunohistochemistry with antibodies against MOG, myelin-associated glycoprotein, glial fibrillary acidic protein, complement C9neoantigen, T cells (CD4, CD8), and macrophages (Kim1p or CD68) as described previously.¹⁶ Apoptotic oligodendrocytes were defined by nuclear condensation and fragmentation in cells stained by either MOG or cyclic nucleotide phosphodiesterase antibodies. For immunopattern classification, all cases were independently reviewed by at least 2 neuropathologically trained experts blinded to the opinion of the other. Consensus opinion was sought in challenging cases. For serial samples from the same patient, each rater was blinded to immunopattern classification rendered on previous tissue samples from the same patient.

Statistical Methods

Survival information for biopsied individuals who were residents of the United States was ascertained from Mayo Clinic institutional databases and a search of the commercial LexisNexis Accurint death records database. All statistical analyses were performed with the patient as the unit of analysis. Even though some patients had biopsy followed by autopsy, they contributed only 1 data point to any analysis, thus maintaining statistical independence among observations. In examinations of the association between immunopattern and biopsy vs autopsy, patients with biopsy followed by autopsy were grouped with other patients with autopsy. For assessing immunopattern vs clinical characteristics, all available data were used, even if not complete for clinical information, to reduce selection bias. We used χ² tests, rank-based Kruskal-Wallis tests, and logistic regression to assess associations. Standard Kaplan-Meier and Cox regression methods were used to assess survival from biopsy. We interpreted p values as goodness-of-fit measures indicating how consistent observed data were with the null hypothesis, recognizing that hypothesis testing depends on an idealized set of assumptions such as random sampling from a population.

Data Availability

Anonymized data and documentation from this study can be made available to qualified investigators on reasonable request. Such arrangements are subject to standard data-sharing agreements.

Results

Neuropathology

A total of 547 individuals fulfilled the neuropathologic diagnostic criteria of inflammatory-demyelinating disease consistent with MS and could be immunopattern classified on the basis of the presence of ≥1 early active lesions. Nine patients (2%) underwent biopsy and subsequently came to autopsy; 494 underwent biopsy only (90%), and 44 patients (8%) underwent autopsy only. A subset of 22 individuals were immunopattern classified at 2 different time points (serial biopsy n = 17, biopsy-autopsy n = 5) and demonstrated the same immunopattern detected at both time points. Furthermore, no patient had >1 immunopattern identified despite possibly having several active lesions assessed, confirming findings of prior studies.^5,11 The cohort had varying degrees of clinical information available, ranging from minimal information such as age and sex to detailed clinical histories with neurologic examination. The median duration from first symptom to biopsy or autopsy (whichever was earliest) was 45 days (range 0 days–32 years). Median time from first attack to last follow-up was 2.3 years (range 0–38 years). A summary of demographics is shown in the Table.

Table.

Patient Characteristics Overall and by Immunopattern

Overall Frequency of Different Patterns of Demyelination in Relation to Clinical Disease at Biopsy or Autopsy

The distribution of the immunopatterns is summarized in Figure 2 and eFigure 1 (doi.org/10.5061/dryad.5x69p8d2t), with pattern II lesions predominating (56%) and roughly equal proportions of pattern I (23%) and pattern III (22%) lesions. The overall pattern frequency was similar to that reported previously.^3,6 Although primarily a cohort of biopsied cases, the smaller subset with autopsy had a markedly different distribution of immunopatterns (p < 0.001, Figure 2). Autopsied individuals were more likely than biopsied individuals to have pattern III (45% vs 19%), much less likely to have pattern I (8% vs 24%), and somewhat less likely to have pattern II (47% vs 57%). Although the majority of patients had a biopsy or autopsy early in their disease course, active demyelinating lesions demonstrating all patterns were detected in 8 patients with disease duration of ≥20 years (Figure 3), albeit at a lower frequency later in the disease course, in agreement with our group's previous findings.⁸ The age at biopsy or autopsy ranged between 4 and 83 years (Table and eFigure 2 [doi.org/10.5061/dryad.5x69p8d2t]), demonstrating that active lesions can occur at any age. Age was not significantly different between the immunopatterns (p = 0.16), nor was there any association between sex and immunopattern (p = 0.71), with women making up 54% of pattern I cases, 56% of pattern II cases, and 51% of pattern III patients.

Figure 2. Distribution of Immunopattern in Overall, Biopsy, and Autopsy Subjects.

Patients coming to autopsy were more likely to be pattern III and less likely to be pattern I or II compared to those who were biopsied (p < 0.001). Autopsy patients in this figure include 9 patients who had previously had a biopsy. These patients were not included in the biopsy group.

Figure 3. Boxplots of Duration From First Attack/Index to Autopsy/Biopsy by Immunopattern.

The distribution of each immunopattern by time from disease onset to autopsy or biopsy is shown here using a transformed x-axis to compress longer disease durations. Patients coming to biopsy are represented by circles. Patients coming to autopsy are represented by triangles. This figure demonstrates that all immunopatterns can be detected more than 20 years after the initial attack, albeit at a lower frequency than closer to the initial attack.

Immunopattern III Was Associated With a Higher Attack-Related Disability

In addition to being overrepresented among autopsy cases, immunopattern III cases had more severe attack-related disability than pattern II cases (median Expanded Disability Status Scale score 4 vs 3, p = 0.02). Despite this, in patients followed up for ≥5 years, there was no difference in the rate of developing progressive disease between the immunopatterns (p = 0.49, Figure 4), and the proportion of patients in the different clinical course categories was similar between the groups.

Figure 4. Recent Clinical Course by Duration From First Attack/Index to Last Follow-up.

These bar charts demonstrate the relative distribution of patients with progressive multiple sclerosis in epochs of disease duration less than 5 years and disease duration greater than 5 years from first attack to last follow-up. There is no difference in the rate of progressive disease between immunopatterns in either epoch (p ≥ 0.49). Dark bars represent patients with progressive MS, lighter bars represent patients without progressive MS.

Immunopattern II Was Overrepresented in Patients With Longer Disease Duration

Among the subset of individuals with disease duration of >5 years at autopsy or biopsy (37 patients), 27 of 37 had immunopattern II, with only 3 of 37 having pattern I and 7 of 37 having pattern III (p = 0.02, Figure 5). Although pattern III cases were less likely to come to biopsy, long-term survival after biopsy was similar (p = 0.64) across the immunopatterns, with ≈90% of individuals surviving 5 years and 80% surviving 10 years after biopsy (Figure 6, which includes the Kaplan-Meier curve demonstrating similar survival for patients who came to biopsy, regardless of immunopattern). Note that disease course at last follow-up was not available for 247 of 547 patients.

Figure 5. Prevalence of Immunopattern With Respect to Disease Duration.

Barplots showing distribution of immunopattern by range of duration from first attack to autopsy or biopsy for patients with a disease duration of less than or equal to 5 years, or greater than 5 years. Patients with a disease duration of more than 5 years tended to have a greater likelihood of having pattern II immunopattern and a greatly reduced likelihood of pattern I immunopattern compared to patients with a disease duration of less than or equal to 5 years (p = 0.02).

Figure 6. Survival From Biopsy Grouped by Immunopattern.

Kaplan Meier curve demonstrating similar survival for patients who came to biopsy irrespective of immunopattern.

Discussion

This work demonstrates that all 3 immunopatterns can be detected in active demyelinating lesions throughout the course of the disease, although with a lower frequency later into the disease course due to the paucity of active demyelinating lesions observed at this stage of the disease. Pattern III appears to be associated with a more fulminant initial attack than patterns I or II, with an overrepresentation of patients with pattern III in the autopsy group. Biopsied patients, a group more likely to have survived the initial attack, appear to have similar long-term outcomes regardless of their immunopatterns. One in 5 patients in our cohort had monophasic disease initially (clinically isolated syndrome) and did not fulfill a clinical definition of relapsing-remitting MS. There was no pathologic difference between these patients and patients who fulfilled the relapsing-remitting MS definition. The majority of these patients converted to clinically definite MS at a long-term follow-up. This supports the hypothesis that clinically isolated syndrome is part of the continuum of MS and should be treated as such.

Heterogeneous immunopatterns have been detected in the early active plaques of patients with MS. Our group has shown that each patient expresses only a single immunopattern across multiple active lesions either serially biopsied or within a single autopsy.⁴ With a 6-fold increase in the number of immunopattern-classified individuals over the last 18 years, we have found that this classification accurately describes the observed pathologic variation in active MS.^3,8 The relative frequency of patients with active plaques, a prerequisite for immunopattern classification, reduces with longer disease duration.¹¹ Active plaques are considered the pathologic substrates of clinical relapses. Therefore, not surprisingly, they are more frequent in acute MS, relapsing-remitting MS, and secondary progressive MS with attacks and are less common in primary progressive MS or secondary progressive MS without clinical attacks. It is worth noting that MS-related progression of the disease is not driven exclusively by the occurrence of relapses.

Prior studies evaluating plaque heterogeneity late in the disease course were likely biased by the lack of active lesions in patients with long-standing progressive disease.¹⁸ The VU Medical Center group¹⁸ evaluated 39 patients with long-standing MS and determined that there was no heterogeneity of the MS plaque between these patients. In fact, the probability of finding a single active demyelinating lesion at MS autopsy after 30 years of disease duration is <1%.⁸ In our cohort, patients with type III immunopattern were overrepresented at autopsy, suggesting that pattern III may be more common in patients with more severe acute attacks of MS, resulting in death and subsequent autopsy. Supporting this is the fact that patients with immunopattern III were more likely to be severely disabled after their initial attack compared to those with pattern II. This excess in disability appears to be front-loaded, with no difference between the immunopatterns in the development of progressive neurologic dysfunction in patients who are followed up for ≥5 years. The underlying pathogenesis of the initial immunopattern appears to predominate in early MS, when active white matter plaques are more common and represent the pathologic substrate of the clinical relapse. Progressive disease appears to be less associated with the initial immunopattern driving the white matter plaque and suggests the potential convergence into a final common pathogenic pathway related to the sequelae of the chronically denuded axon.

The pathology of MS-related progression is complex, with neurodegeneration amplified by accumulation of CNS damage and aging.¹⁹ In addition to focal active plaques, smoldering white matter plaques are found only among patients with progressive MS.⁸ Multiple pathologies underlie disease progression, including axonal degeneration, remyelination failure, cortical demyelination, and meningeal inflammation. Once the axon is deprived of the trophic support from myelin and oligodendrocytes, the increasing energy demand results in upregulation of sodium channels along the demyelinated axon. Although aiding in the conduction of action potential, it is not an efficient strategy and requires more energy.²⁰ To meet these increased energy demands, axonal mitochondria are recruited to areas of demyelination; however, over time, the antioxidative mechanisms fail, and oxidative damage ultimately leads to mitochondrial dysfunction,²¹ which translates into irreversible axonal damage. This final common pathway, with mitochondrial loss and axonal death, is likely unrelated to the initial immunopattern causing demyelination.

Developing a noninvasive (most promising radiologic) biomarker of immunopatterns is needed to better determine the clinical relevance of the immunopattern and to facilitate individualized treatment on the basis of the underlying pathology. Given that pattern III lesions represent an oligodendrocytopathy and pattern II lesions are more suggestive of humoral immunity, vastly different therapeutic approaches are likely to be beneficial in each case. It already has been shown with regard to relapse treatment: patients with pattern II lesions benefited most from plasma exchange, while patients with pattern I or III lesions were not likely to respond.^22,23 To enable targeted therapy in this fashion, a noninvasive or minimally invasive biomarker would be necessary to identify the immunopattern in individual patients. Our data show that such a biomarker would be most useful around the time of the development of the active lesion.

This study is limited by the potential for selection bias: patients with MS who come to biopsy or autopsy may not be representative of the general MS population. Our ongoing evaluations of the representativeness of our cohort indicate that despite an atypical presentation, biopsy cases have a typical clinical course and provide a valid classification of the neuropathology found in patients with MS in general.⁵ While the distribution of the immunopatterns within the general population may differ from what we have reported, we believe that we have representative samples of each pattern in MS and that comparisons remain valid. Still, it is likely that type II immunopattern predominates, given that it was the most common in the biopsy cohort.

The overrepresentation of pattern III in patients who came to autopsy suggests that the acute disability associated with the initial attack is more severe in patients with pattern III disease.^4,24 Because patients in this study were not followed up at regular intervals during the course of their disease, we were unable to say at what point they developed progressive MS. Despite this limitation, we were able to evaluate the ultimate outcome from the point of progression vs immunopattern and to assess long-term survival, and there did not appear to be a difference between the groups. Further work is needed to identify noninvasive biomarkers of MS pathology, including the use of MRI markers. We have not presented MRI findings in this study; they will be the subject of a future study.

This study provides insights into the clinical relevance of heterogeneous immunopatterns detected in the early active plaque of patients with MS. The differential outcome from the acute attack suggests that tailored initial treatment from the point of view of either attack recovery or attack prevention enhances attack recovery or prevents the onset of secondary progressive disease.

Glossary

MOG

myelin oligodendrocyte glycoprotein

MS

multiple sclerosis

Appendix. Authors

Study Funding

This study was funded by Novartis (CFTY720DUS37T) and the NIH (R01NS49577-7). J.M.F. was supported by the Austrian Science Fund (FWF Project J3508-B24) while conducting parts of this work. W.O.T. receives research funding from the Mayo Clinic Center for MS and Autoimmune Neurology.

Disclosure

W. O. Tobin receives research funding from the Mayo Clinic Center for MS and Autoimmune Neurology. A. Kalinowska-Lyszczarz, S. Weigand, Y. Guo, N. Tosakulwong, J. Parisi, and I. Metz report no disclosures relevant to this manuscript. J. Frischer was supported by the Austrian Science Fund (FWF Project J3508-B24) while conducting parts of this work. H. Lassmann, W. Brück, L. Linbo, and C. Lucchinetti report no disclosures relevant to this manuscript. Go to Neurology.org/N for full disclosures.