J H Simon and BK KleinschmidtDeMastersthese terms to denote cert - PDF document

1 J. H. Simon and B.-K. Kleinschmidt-DeMas
J. H. Simon and B.-K. Kleinschmidt-DeMastersthese terms to denote certain stereotypic sites of involvement irrespective of co-existent disorders. An example of this would be the use of Devics diseaseŽ for spinal cord and optic nerve lesions in autoimmune disorders or cancer.Today, based on neuropathology and imaging, Marburg (acute MS), Schilder (large coalescent, usually childhood, MS with exclusive cerebral hemispheric involvement), and Balo (alternating concentric rings of myelin loss and preservation in MS) are accepted as variants of MS. Devics disease ned by exclusive spinal cord and optic nerve involvement) on the other hand, from both an im-aging and neuropathology perspective is less con“dently categorized today as an MS variant, but if it is, it likely lies on the far-side of the spectrum of MS (Table 16.1).All these variants result in diagnostic dif“ culty and errors disproportionate to their incidence. A few additional points are worth noting:1) The initial clinical presentation of these MS variants is usually severe and rapidly progres-sive, making the differential diagnostic list broad.2) Making the diagnosis at the very initial phases of any of these diseases (when the neuroradiologist is most likely to be confronted with the disorder by imaging studies) is virtually impossible. Only passage of time and evolution of clinical features of the disease, biopsy, or even autopsy provide the nal and correct diagnosis.3) All these MS variants are rare and heteroge-neous within each category due to differing uses of the terms, as noted above. Although correct diagnosis is always desirable, in actuality, treat-ment decisions are more often guided by anec-dote and desperation than by evidence-based 4) Despite these dif“ culties, from a research perspec-tive, these MS variants as outliersŽ do provide impor

2 tant material with which to address ques
tant material with which to address ques-tions related to variant host response, mechanisms of remyelination or barriers to remyelination, and issues surrounding new hypotheses of MS classi“cation (Lucchinetti et al. 2000).This chapter will attempt to outline the unique versus the overlapping clinical, pathological, and neuroimaging features of Marburg, Schilder, Balo, and Devic types of MS. Acute tumefactive demyelin-ating lesions will also be brie” y discussed under the acute MS section.Devics Neuromyelitis OpticaGeneral FeaturesDevics neuromyelitis optica (DNO), originally de-scribed in 1894 (Devic ammatory de-myelinating disease with features that overlap with MS Cree et al. 2002; et al. 2003; Weinshenker2003). However, in contrast to the other MS variants discussed below, DNO has come to be understood as more distinct than similar to MS, based on clinical, pathological, immunological and imaging criteria.The key clinical features of DNO include acute vi-sual loss, often bilateral, and acute transverse myelitis, with the visual and spinal cord signs and symptoms often presenting nearly simultaneously (ORiordanet al. 1996; et al. 2003; Wingerchuk et al. 1999). Despite striking and classic demyelination in the optic nerve and spinal cord, patients with pureŽ DNO do not develop neurologic signs or symptoms or demyelination in other regions of the CNS; the brain remains normal or shows at most non-speci“ c “ nd-ings in the white matter on follow-up MRI (Filippi et al. 1999; et al. 1993; ORiordan et al. 1996; Fazekas et al. 1994).Many series of DNO patients suggest that visual symptoms usually precede spinal symptoms, but the reverse is not uncommon. Severe visual symptoms with blindness may become total and permanent within a few days and bilateral optic neuritis is most common. This contrasts with MS where

3 bilateral op-tic neuritis is relatively
bilateral op-tic neuritis is relatively uncommon and the initial visual compromise is usually limited and reversible Weinshenker 2003). In patients who present with visual system dif“ culties, transverse myelitis charac-teristically develops within a few weeks, with severe paraplegia, sensory loss with a distinct level, and sphincter disturbances. Fixed weakness from onset, rather than improvement over time after the initial event, is the typical course. This contrasts with MS where weakness is frequently reversible in the early stages and the spinal cord presentation is that of only partial transverse myelitis. The interval separat-ing the visual and spinal syndromes may be hours, days, or weeks, but in most cases is within 3 months. However, cases with a 2- or more year separation have been described. As no single criterion is speci“ c for DNO, diagnostic criteria have been developed based on multiple factors, including MRI (WingerchukWeinshenker 2003; Wingerchuk et al. 1999; et al. 2002). Variants of Multiple Sclerosis While there is some variation depending on the criteria used to make the diagnosis of DNO, there is good evidence that the majority of cases will be characterized by a normal brain MRI at presenta-tion (Wingerchuk et al. 1999; Wingerchuk and Weinshenker 2003; et al. 1993; et al. 2003; ORiordan et al. 1996; Fazekas et al. 1994; Filippi et al. 1999). When T2-hyperintensities are seen in the brain, they tend to be non-speci“ c, for example not abutting the ventricular surfaces, and without accompanying T1-hypointensity, either acute or chronic. On follow-up, cases of pureŽ DNO do not tend to accumulate new T2-hyperintensities, in con-trast to typical MS where new lesions are frequent and most often subclinical (Filippi et al. 1999).Magnetization transfer imaging studies have provided further sup

4 port for the concept of DNO as a more li
port for the concept of DNO as a more limited neuroanatomical disorder in con-trast to typical MS. The magnetization transfer ra-tio (MTR) of focal non-speci“ c brain lesions when they do occur in DNO is only mildly abnormal and considerably less abnormal than in MS. The MTR-based measures of normal-appearing white mat-ter (NAWM) is normal in DNO, in contrast to the NAWM in MS which is consistently abnormal even in the relatively early stages of disease (Filippi et al. 1999). In keeping with a more severe spinal cord pathology, the MTR of spinal cord is signi“ cantly more abnormal in DNO than in typical MS (Filippiet al. 1999). b Fig. 16.1a-f. Devics neuromyelitis optica. A relapsing spinal cord presentation included an episode with severe long segment, diffuse spinal cord involvement as seen on T2-weighted MRI (), with corresponding T1-hypointensity (), “ ndings not as-sociated with classic MS. On axial T2-weighted MRI (), the cervical cord lesion is large, symmetric, and crosses the midline. The brain MRI in Devics neuromyelitis optica is typically normal as shown here (), but can include non-speci“ c T2-hyperin-tensities. In MS, on T2-weighted sagittal images, the spinal cord lesion is often vertically oriented, but ments in height. Involvement as seen on axial T2-weighted axial images () is classically asymmetric, corresponding to the clinical “ nding of partial transverse myelitis J. H. Simon and B.-K. Kleinschmidt-DeMasters erential DiagnosisDepending on the initial presentation (spinal or op-tic), the differential diagnosis by imaging includes (for spinal presentations) transverse myelitis, acute disseminated encephalomyelitis (ADEM), including post-vaccination, viral and other infectious etiologies, and neoplasm. An abnormal brain MRI with typical demyelinating lesions in the cerebral hemispheres, cerebel

5 lum, and brainstem, and consistent labor
lum, and brainstem, and consistent laboratory ndings makes MS more likely. Vascular insults and collagen vascular disease are not always separable from DNO without supporting laboratory or histori-cal information. With an optic neuritis presentation, especially if unilateral, imaging “ ndings are non-spe- c and the differential diagnosis will include optic neuritis from MS if the brain MRI is positive. Acute MS (Marburg Type)General FeaturesRarely, an acute, idiopathic, in” ammatory demy-elinating disease may be relatively unresponsive to conventional therapy (corticosteroids), resulting in death or severe residual de“ cits, and may then be ed as acute MS of the Marburg type. Death may occur in weeks to months, either from severe widespread cerebral lesions or acute involvement of the lower brainstem or upper cervical cord (Pogacar 1988). Patients who survive the acute presentation may be left with signi“ cant de“ cits or may develop severe exacerbations.It must be acknowledged, however, that some pa-tients who present with a large acute tumefactive de-myelinative lesion that prompts biopsy, later go on to exhibit a classic chronic relapsing/remitting course of typical MS. These cases should perhaps not be con-sidered pureŽ Marburg type MS although at the time the patient presents the outcome may not be at all clear. For the individual patient, the dif“ culty is that at the time of clinical presentation there are no good predictors for whether he or she will follow a fulmi-nant rapidly fatal course after biopsy, develop mild or severe MS, or even develop MS at all, despite several years follow-up (Kepes 1993). It has been suggested that some patients who present with acute tumefac-tive demyelinating lesions prompting biopsy might actually have disorders intermediate between MS and large coalescent ADEM (Kepes 19

6 93). Hence, the appellation MarburgŽ is
93). Hence, the appellation MarburgŽ is best applied to severe, acute MS that meets both clinical and pathological criteria of a monophasic illness and may best be de“ ned by its malignant course and acute, severe demyelination et al. 1999; Poser et al. 1992).While poorly responsive to corticosteroids (the “ rst line of therapy), there are suggestions that patients with Marburg-type MS may bene“ t from plasma ex-Weinshenker 1999). Cases have also been described with response to combined corticosteroid and mannitol therapy (Giubilei et al. 1997).Recently, Marburg MS has been hypothesized to be the result of a pre-existing abnormality of myelin basic protein in a developmentally immature (less cationic) form (Wood et al. 1994; Beniac et al. 1999). Alternatively, Marburg type of demyelination may simply lie at the severe, acute end of the clinical spec-trum of MS (Coyle 2000), possibly re” ecting factors related to a host response.NeuropathologyIn most patients who succumb, demyelinative lesions are usually all of the same acute age. The distribution of lesions is similar to typical MS, but there is a predi-lection for lesions to occur in the cerebral hemispheres, including near the gray-white matter junction. Virtually no older plaques can be identi“ ed at neuropathologi-cal examination, paralleling the patients rapid clinical downhill course. This is in contradistinction to typical MS where the demyelinative lesions at autopsy are usu-ally remote and/or show differing ages of myelin break-down. The latter is assessed by several types of special stains, the most common of which is the histochemical stain for myelin (Luxol fast blue, LFB) with a periodic acid Schiff (PAS) counterstain. This stain allows distinc-tion between very recently phagocytosed myelin within macrophages, which is as yet undigested and retains

7 its LFB-positivity, versus myelin that
its LFB-positivity, versus myelin that has been broken down to neutral lipids in macrophages (PAS-positivity) as a result of a more advanced and subacute process. In Marburg type of acute MS, abundant LFB-positive macrophages are seen throughout the hypercellular de-myelinative lesions. Acute MS shows less chronic glio-sis, more edema, and even partial bands of preserved myelin (see description of Balo type in Sect. 16.4) than typical relapsing/remitting MS cases that come to au-topsy. More abundant perivascular in” ammation may also be present than is seen in typical MS. J. H. Simon and B.-K. Kleinschmidt-DeMastersThe explanation for the pre-MRI and MRI era dichotomy is not known. One possibility may lie with a more detailed explanation of the pathological features of acute MS. Acute Marburg type MS often shows partial bands of preserved myelin at patho-logical examination, as explained above. Such cases with limited banding are classi“ ed as acute MS by neuropathologists, but might be considered as hav-ing focal or limited Balo ringsŽ based on imaging. Acute tumefactive demyelinative lesions with a few alternating bands of intact and damaged myelin may be the lesions that account for the diagnosis of Balo ringsŽ by MRI in some patients.Co-existence of BCS lesions with typical MS-like (non-concentric) lesions has also been recognized in the neuropathology literature (Itoyama et al. 1985; Moore et al. 1985). BCS lesions have been found very rarely to develop after a typical relapsing-remit-ting course of typical MS (Moore et al. 2001), and Balo-like band pattern lesions have been described along the periphery of acute MS plaques (Moore et al. 2001).While the early literature emphasized BCS within the cerebral hemispheric white matter of the brain, this process also occurs within the optic chiasm and spinal cord,

8 and BCS lesions can occur in the brain-
and BCS lesions can occur in the brain-stem and cerebellum (Itoyama et al. 1985; Mooreal. 1985). NeuropathologyWhile much of the literature of BCS was interpreted as indicating that the abnormal bands (intermingled with normal bands) were the result of partial re-myelination, more recent evidence suggests that the abnormal bands are more often areas of early demy-elination alternating with preserved myelin (Yao et al. 1994). In some well documented cases there does appear to be remyelination of previously demyelin-ated “ bers (Moore et al. 1985). BCS-like features, without the full blown BCS macroscopic pathology have also been described along the surface of chronic active MS plaques (Moore et al. 2001), suggesting that this pathology may not be characteristic of a separate disease, but re” ects a variation in pathology or host response.The mechanism(s) responsible for this peculiar pathology remain a mystery. Balo postulated a leci-thinolytic enzymeŽ centrifugally spreading from the center of a lesion (Kuroiwa 1985). A more recent hypothesis is based on downregulation of local de-myelination by CD8 suppressor T-cells (Moore et al. 2001; Traugott et al. 1983), in both typical MS and BCS lesions, as might occur as well through cytokine action (Moore et al. 2001; Canella and Raine 1995). The suppression of demyelination along the surface of a lesion, surrounded by an external zone of activ-ity and demyelination, and a sequential repeat of the process might then result in the repetitive alternating lamellae characteristic of BCS (Moore et al. 2001).With the introduction of MRI, the literature regard-ing BCS has evolved. BCS is currently described as consisting of a range appearances, from classic large BCS lesions in isolation associated with a fulminant clinical course, to cases in which the BCS pattern of focal le

9 sions co-exists with typical MS-like les
sions co-exists with typical MS-like lesions et al. 1996; Yao et al. 1994; Iannucci et al. 2000; et al. 1999). There is also increasing recog-nition of borderline BCS-like lesions with only a few lamellae or rings. The co-existence and development of BCS lesions within typical MS lesions, though rare, is likely more common than development of only classic large BCS lesions throughout both cerebral hemispheres.BCS lesions are readily identi“ ed on proton or T2-weighted images, but the concentric pattern may also be apparent on T1-weighted images (Fig. 16.3). There are reports of contrast enhancement in BCS, with al-ternating bands of enhancement and non-enhance-ment, the enhancing regions thought to correspond to zones of demyelination. Synchronously enhanc-ing, sequentially enhancing, and transiently enhanc-ing rings have been reported (Iannucci et al. 2000; Sekijima et al. 1997; 2001; et al. 1999; et al. 1999; Bolay et al. 1996; Caracciolo et al. 2001; Kastrup et al. 2002) (Fig. 16.4).By both imaging and neuropathology, most studies suggest a chronologic progression of rings, with the most recent pathology along the periphery. In some cases a more synchronous process has been sug-gested, possibly associated with a very rapid tempo of lesion progression, not unlike that of chronic active lesions in classic MS, where ongoing demyelination may occur along a lesions perimeter (Moore et al. 2001; Sekijima et al. 1997; et al. 1999).In the few cases of BCS studied to date with MR spectroscopy, the principal metabolite ratios and other abnormal peaks (from lipid and lactate) are those observed from typical large MS lesions J. H. Simon and B.-K. Kleinschmidt-DeMasters erential DiagnosisClassic large and potentially mass-like BCS lesions may be confused with neoplasm or abscess, but when their appearance is equivocal,

10 the course when followed by imaging sho
the course when followed by imaging should suggest a resolving (non-malignant) process. MR spectroscopy may also support the diag-nosis of a demyelinating process ( et al. 2001).rings representative of BCS (Fig. 16.5), and its corre-sponding pathology, or are these better characterized as relatively rare presentations of the common demy-elination characteristic of typical MS? Does identi“cation of a minimal BCS pattern contain prognostic information, information relevant to discussions of subtypes of MS pathology or information relevant to optimal therapy?Schilders DiseaseGeneral FeaturesSchilders disease, also known as diffuse myelinocla-stic sclerosis, is a rare demyelinating disorder that is essentially a diagnosis of exclusion as its clinical and MRI appearance may overlap with inherited metabolic disorders of myelin, particularly adrenoleukodystro-phy. Schilders original three cases, under the umbrella of an entity known as encephalitis periaxialis diffusa, described in 1912, 1913, and 1924 were subsequently found to be three different disorders: myelinoclastic diffuse sclerosis (the 1912 case), adrenoleukodystro-phy (the 1913 case), and subacute sclerosing panen-cephalitis (SSPE) (the 1924 case) (Poser 1985).A practical de“ nition proposed by Poser (1985) includes the following components: (1) a subacute or chronic myelinoclastic disorder with one or two roughly symmetrical plaques at least 2×3 cm in two of three dimensions; (2) involvement of the centrum semiovale; (3) these being the only lesions based on clinical, paraclinical or imaging “ ndings; (4) adreno-leukodystrophy must be excluded.Schilders disease in its classic sense is acute MS that occurs in childhood. Analogous to other MS vari-ants discussed above, both pureŽ forms and transi-tionalŽ forms have been described. PureŽ forms pre-dominat

11 e in childhood and have plaques con“ ned
e in childhood and have plaques con“ ned to the cerebral white matter. TransitionalŽ forms af-fected a broader age range of adolescents and adults and show large cerebral plaques combined with more typical MS plaques elsewhere (Prineas et al. 2002).Disease duration is highly variable. In the 70 cases collected by Poser (1957) the mean duration was 6.2 years, ranging from 3 days to 45 years, but dura-tion was less than 1 year in 40%. The clinical course is diverse, but widespread white matter involvement usually produces subacute or chronic mental and neurological deterioration, spastic paresis, convul-sion, and involvement of vision and hearing. Pure psychiatric forms have been described. Increased intracranial pressure, headache, and vomiting may suggest a mass lesion. The cerebrospinal ” uid may be normal with only a slight elevation of protein or cell count. Treatment with corticosteroids and cyclo-phosphamide has been successful in some instances. Cases considered to be Schilders disease at onset may follow a downhill progressive course or go on to de-velop relapsing/remitting disease. As noted above in the discussion of acute MS, it is not possible to deter-Fig. 16.5a,b. Balos concentric sclerosis pattern with few rings. Biopsy-proven in” ammatory demyelinating process consis-tent with Balos concentric sclerosis. T2-weighted image (shows two concentric zones of abnormal intensity. Post-con-trast T1-weighted image () shows multiple concentric rings of enhancement. [From Caracciolo et al. (2001)] a b Variants of Multiple Sclerosis Green“ elds neuropathology, 7th edn. John Wiley & Sons, New YorkSchwarz S, Mohr A, Knauth M et al (2001) Acute disseminated encephalomyelitis: a follow-up study of 40 adult patients. Neurology 56:1313-1318Sekijima Y, Tokuda T, Hashimoto T et al (1997) Serial magnetic resonance ima

12 ging (MRI) study of a patient with Balo
ging (MRI) study of a patient with Balos concentric sclerosis treated with immunoabsorption plas-mapheresis. Mult Scler 2:291-294Simon JH (2000) Brain and spinal cord atrophy in multiple sclerosis. Neuroimaging Clin N Am 10:753-770Spiegel M, Kruger H, Hofmann E et al (1989) MRI study of Balos concentric sclerosis before and after immunosup-pressive therapy. J Neurol 236:487-488Traugott U, Reinherz EL, Raine CS (1983) Multiple sclero-sis. Distribution of T cells, T cell subsets and Ia-positive macrophages in lesions of different ages. J Neuroimmunol Valk J, van der Knaap MS (1995) Multiple sclerosis, neuro-myelitis optica, concentric sclerosis, and Schilders diffuse sclerosis. In: Magnetic resonance of myelin, myelination, and myelin disorders. Springer-Verlag, Berlin-Heidelberg-New York, pp 179-205Weinshenker BG (2003) Neuromyelitis optica: what it is and what it might be. Lancet 361:889-890Weinshenker BG (1999) Therapeutic plasma exchange for acute in” ammatory demyelinating syndromes of the cen-tral nervous system. J Clin Apheresis 14:144-148Wingerchuk DM, Hogancamp WF, OBrien PC et al (1999) The clinical course of neuromyelitis optica (Devics syndrome). Neurology 53:1107-1114Wingerchuk DM, Weinshenker BG (2003) Neuromyelitis optica: clinical predictors of a relapsing course and sur-vival. Neurology 60:848-853Wood DD, Bilbao JM, OConnors P (1994) Acute multiple scle-rosis (Marburg type) is associated with developmentally immature myelin basic protein. Ann Neurol 40:18-24Yao DL, Webster HD, Hudson LD et al (1994) Concentric scle-rosis (Balo): morphometric and in situ hybridization study of lesions in six patients. Ann Neurol 35:18-30for humoral mechanisms in the pathogenesis of Devics neuromyelitis optica. Brain 125:1450-1461Lycklama G, Thompson A, Filippi M et al(2003) Spinal-cord MRI in multiple sclerosis

13 . Lancet Neurol 2:555-562Mandler RN, Dav
. Lancet Neurol 2:555-562Mandler RN, Davis LE, Jeffery DR et al (1993) Devics neuro-myelitis optica: a clinicopathological study of 8 patients. Ann Neurol 34:162-168Mandler RN, Ahmed W, Dencoff JE (1998) Devics neuromyeli-tis optica: a prospective study of seven patients treated with prednisone and azathioprine. Neurology 51:1219-1220Mendez MF, Pogacar S (1988) Malignant monophasic multiple sclerosis or Marburgs diseaseŽ. Neurology 38:1153-1155Mehler MF, Rabinowich L (1988) In” ammatory myelinoclastic diffuse sclerosis. Ann Neurol 23:413-415Moore GR, Neumann PE, Suzuki K et al (1985) Balos con-centric sclerosis: new observations on lesion development. Ann Neurol 17:604-611Moore GR, Berry K, Oger JJ et al (2001) Balos concentric scle-rosis: surviving normal myelin in a patient with a relaps-ing-remitting clinical course. Mult Scler 7:375-382Noseworthy JH, Lucchinetti C, Rodriguez M et al (2000) Mul-tiple sclerosis. N Eng J Med 343:938-952Nejat F, Eftekhar B (2002) Decompressive aspiration in myeli-noclastic diffuse sclerosis or Schilder disease. Case report. J Neurosurg 97:1447-1449Ng SH, Ko SF, Cheung YC et al (1999) MRI features of Balos concentric sclerosis. Br J Radiol 72:400-403ORiordan JI, Gallagher HL, Thompson AJ et al (1996) Clinical, CSF, and MRI “ ndings in Devics neuromyelitis optica. J Neurol Neurosurg Psychiatry 60:382-387Poser CM (1957) Diffuse-disseminated sclerosis in the adult. J Neuropathol Exp Neurol 16:61-78Poser CM (1985) Myelinoclastic diffuse sclerosis. In: Koetsier JC (ed) Demyelinating diseases. Handbook of clinical neu-rology, vol 3, pp 419-428Poser S, Luer W, Bruhn H et al (1992) Acute demyelinating disease. Classi“ cation and non-invasive diagnosis. Acta Neurol Scand 86:579-585Prineas JW, McDonald WI, Franklin RJM (2002) Demyelinat-ing diseases. In: Adams JH, Corsellis JAN, Duch