Multiple sclerosis: Classification revisited reveals homogeneity and recapitulation

Abstract

Classifying the multiple sclerosis (MS) lesion, the subject of this editorial highlighting Breij and colleagues'1 article in this issue of Annals, has proved something of a contentious topic over recent years. Not entirely surprising for a disease such as MS with its wide-ranging clinical and pathological phenotype, a single predominant disease mechanism has not yet emerged. Although few would query the compelling evidence for an immune-mediated pathogenesis,2 precisely how the various players of the immune system contribute to the pathognomonic established MS plaque remains an enigma. It is widely held that the initial lesion in MS follows a systemic infection that causes proinflammatory cytokines to be released into the general circulation, leading to upregulation of adhesion-related molecules on central nervous system (CNS) endothelium and homing of leukocytes to CNS vasculature, which traverse vessels to enter the normally sequestered CNS. Should lymphocytes programmed to recognize myelin antigens be present in the infiltrate, then the stage is set for the formation of an acute, inflammatory demyelinating lesion. This hypothetical scenario, supported by immunopathological reports on MS plus convincing data from animal models, is believed to underlie the MS plaque, which has traditionally and didactically been presented as developing in a spectrum of acute and chronic stages, matched clinically by exacerbations and remissions. However, recognizing the current consensus that lesion pathogenesis in MS may be heterogeneous, any proposal suggesting that lesions develop in a uniform, linear fashion is likely to invite merited criticism. In the late 1990s, followers of the above spectrum concept were challenged by a series of insightful communications from the laboratories of Lucchinetti and colleagues, who took on the formidable task of rearranging MS lesions according to perceived underlying pathological mechanisms, as opposed to stages. The beginning reclassification strategy centered not on immunopathogenesis per se but on the response of the myelinating cell, the oligodendrocyte. In their first article, published in 1996, these investigators were able to discriminate among five lesional types based on oligodendrocyte loss.3 In 1998, there followed one review further analyzing these five lesion subtypes and the associated oligodendrocyte pathology,4 and a second review that mentioned “at least five different subtypes,”5 implying the possibility of more. Then in 1999, a careful morphometric analysis of oligodendrocytes in lesions from 113 cases of MS appeared in which two lesion groups (further divided into five subtypes) were identified: one showing presence and the other absence of oligodendrocyte recruitment.6 Retrospectively, and given the enormity of the task, the authors deserve considerable credit for even attempting to bring order to a lesion as complex as the MS plaque. Nevertheless, the above classification systems appear not to have served their purpose because in 2000, the same authors, approaching the problem from a different standpoint, presented the heterogeneic MS lesion as a collection of lesion types defined not by oligodendrocyte-related changes as before, but by distinct pathogenetic mechanisms.7 In this much-quoted, important article, 4 lesion types (patterns) were identified from paraffin blocks from 81 cases of MS (32 autopsy, 49 biopsy), displaying short disease duration and active demyelination. Remarkably, lesion patterns were stated to be heterogeneous between patient subgroups but homogeneous in different lesions from the same patient. The four patterns were segregated according to their being mediated by T cells and macrophages alone (pattern I), immunoglobulin (Ig) and complement (pattern II), apoptosis of oligodendrocytes and absence of Ig, complement and remyelination (pattern III), and oligodendrocyte dystrophy and no remyelination (pattern IV). This article marshaled a number of previously published mechanisms deemed relevant to the acute MS lesion. Reiterated in numerous reviews spawned since 2000 (eg, Lassmann and colleagues8-11), the four-pattern system still awaits confirmation; although without highly comparable clinical material, this promises to be difficult. Probably the most one might expect in the foreseeable future will be that selected facets of the four-pattern dogma will be reappraised separately. In this latter regard, this issue of Annals hosts an article by Breij and colleagues,1 a group with a long and respectable track record in MS, who have examined pathogenetic events in inflammatory, actively demyelinating lesions from a series of 27 cases of established MS and as control subjects, inflammatory, nondemyelinating, or inactive lesion areas from another 12 cases. Magnetic resonance imaging–guided sampling of lesions was employed, and the classification strategy and technologies applied previously7 were followed carefully. Additional markers for complement, Fc receptors, human leukocyte antigen D related, fibrinogen, and hypoxia inducible factor-1α were tested, and frozen sections were incorporated for more precise localization of immune system molecules. In contrast with the earlier study on fulminant cases of short duration,7 Breij and colleagues1 investigated to what extent the four-pattern criteria translated to active lesions from patients with established MS (mean duration, 22.2 years). To this reader, Breij and colleagues'1 study appears meticulously executed. Lesion heterogeneity was carefully scrutinized, and several parameters key to the four-pattern system, namely, localization and role of IgG and complement, preferential loss of myelin-associated glycoprotein, occurrence of oligodendrocyte apoptosis, and hypoxia-like damage, were diligently weighed. Made abundantly clear from the current results was that active chronic lesions exhibited consistent association with deposition of Ig and complement on macrophages, a feature equating with pattern II in the previous study.7 Moreover, oligodendrocyte apoptosis, a key determinant for patterns III and IV,7 was absent or rare. Likewise, hypoxia-like damage to oligodendrocytes, a criterion added later to the four-pattern system,12 was rare. More significant were observations that active established MS lesions displayed an homogeneous profile, that lesion heterogeneity could not be confirmed, and that there was no interindividual heterogeneity with respect to Ig and complement immunoreactivity. Thus, according to this report,1 as far as actively demyelinating established lesions are concerned, several building blocks of the dogma for lesion categorization that Lucchinetti and colleagues laid down for early active lesions appear not to hold up. How can we reconcile the differences? Most workers in the field have been raised on the mantra that expanding chronic MS lesions probably recapitulate acute events and would be quite comfortable with many of these observations having relevance to the acute lesion. Certainly many similarities exist between the two, such as the expression of the same immune system molecules and the occurrence of identical patterns of myelin pathology among acute and chronic active MS lesions, for example, IgG on macrophages,13, 14 attachment of (?opsonized) myelin debris to clathrin-coated pits on macrophage surfaces,13, 15 and association of myelin-specific IgG with degenerating myelin.14 Thus, it is probably safe to assume that chronic active lesions recapitulate events encountered in acute lesions. Although only direct comparison between acute and chronic active lesions will settle comments such as whether these patterns represent those of a common later pathway (unlikely in view of the above-mentioned similarities), incontrovertible and intriguing from this detailed analysis was that homogeneity and involvement of Ig and complement were consistent features in active older lesions. In the interest of completeness, it is worth noting that Barnett and Prineas's16 2004 study of acute MS reported lesions containing oligodendrocyte apoptosis, complement activation, and remyelination, findings that suggest overlap in the four-pattern system, leaving these authors to question the validity of pattern III. More recently, Stadelmann and colleagues,12 reporting on a rare variant of acute MS, Baló's concentric sclerosis, seemingly restricted pattern III to the concentric lesions. Because only 8 of the original 22 cases displaying pattern III were Baló's,7 one is left to wonder about the nature of the displaced cases. Finally, there has long been discussion on the significance of the oligodendrocyte response in the active MS lesion. The observed lack of oligodendrocyte apoptosis that Breij and colleagues1 detailed is in accord with previous accounts on active established lesions in MS (eg, Raine and colleagues17 and Bonetti and Raine18) and with descriptions of oligodendrocyte survival, proliferation, and protection,17-20 observations militating against an apoptotic mechanism for the oligodendrocyte depletion in MS. Where then does this leave us with regard to the classification of the active MS lesion? Does this account (and that of Barnett and Prineas16) mean that those MS lesions separated by oligodendrocyte behavior3-6 (systems apparently usurped), as well as patterns III and IV of the 2000 report,7 might be in doubt? In the absence of works confirming the four-pattern system, the negatives currently appear to outweigh the positives. From Breij and colleagues'1 report, lesions did not segregate by oligodendrocyte behavior and had a uniform pattern II phenotype. If recapitulation of immunopathogenetic events in acute and chronic active lesions is valid, then categorization of lesions according to oligodendrocyte pathology may have to be excluded from the schema, bringing us back to a system akin to the spectrum concept that was in vogue before 1996. In the famous last words of Ned Kelly (the Australian “hero” bandit), “Such is life.”