The pathognomonic lesions of MS, calledplaques, were chronic inflammatory foci randomlyaffecting the white matter of the central nervous system that resulted in myelin loss and gliosis F
Trang 1manual dexterity, impaired verbal memory and language deficits in all forms of the disease Corticalaphasia, agnosia, and apraxia are rare in MS, whileverbal fluency and verbal memory are often impairedrelatively early during the disease Callosal discon-nection as well as alexia without agraphia was described in case reports (Mao-Draayer and Panitch,2004) Since the observed cognitive abnormalitiespredominantly affect executive functions, such impair-ments by themselves may become highly disabling
in MS, and significantly interfere with professional andsocial functioning The combination of abnormal-ities in attention, planning, working memory, speed
of information processing and visuo-spatial skills,along with physical disability, can significantlyinterfere with the performance of complex dailytasks Impairments in all cognitive domains mayresult from a diffuse distribution of microscopicpathology, while a large lobar lesion can presentwith a predominant lobar deficit Extensive corticalpathology accompanying varying loads of subcort-ical lesions may result in mixed forms of dementia(Buchanan et al., 2005) The severity of cognitiveimpairment best correlates with the total cerebraldisease burden defined by recently developed con-ventional and nonconventional MRI sequences, andboth gray- and white-matter atrophy contributes tocognitive and neuropsychological impairments in
MS (Sanfilipo et al., 2006) Metabolic and functionalabnormalities detected by PET scan or functionalMRI in cortical neurons likely reflect disruption
of intercortical and subcortical pathways, lesionsdirectly affecting neurons and toxic effect of solubleinflammatory products (De Souza et al., 2002;
La Rocca, 2000; Rao et al., 1991) A trans-synapticalteration of neuronal activity is also possible
Mapping of compensatory changes and plasticity ofthe brain represents an important field of functionalimaging (Tartaglia and Arnold, 2006)
Psychological disability in MS most commonly cludes emotional lability, irritability, euphoria, apathy,depression, bipolar disorder, suicidal ideation, anti-social behavior, and psychosis (Figved et al., 2005;
in-De Souza et al., 2002) These symptoms negativelyinfluence the quality of life and add to the disablingeffects of cognitive abnormalities Depression may
be caused by the disruption of normal anatomy,changes in neurotransmitter production, and altera-tion of the neuroendocrine pathways Reaction todisability and medication side effects may also con-tribute to depression Most studies testing the relation-ship between depression and cognition suggest that
there is little or no relationship However, a analysis by Thronton and Naftail (1997) reveals astrong correlation between depression and workingmemory, but no relationship between depressionand short-term or long-term memory ( La Rocca,2000; De Souza et al., 2002) Euphoria is an inappro-priate expression of optimism and happiness that
meta-is often associated with signs of emotional hibition Euphoria usually results from a diffuse andsevere pathology in patients with advanced physicaland cognitive disability
dysin-Bedside testing cannot adequately assess cognitivefunction or mood disorders, and the use of compre-hensive neuropsychological batteries may be ne-cessary in a great proportion of MS patients Theincreasing availability of immunomodulatory, neuro-protective, antipsychotic, and mood-stabilizer drugs,along with other symptomatic treatments and rehabilitation methods, underscore the importance
of early evaluation of cognitive and mood disorders
in MS
Variants of MS
ON, ATM, Marburg’s type of MS and Balo’s centric sclerosis are discussed above Neuromyelitisoptica or Dévic’s disease is reviewed in Chapter 4
syn-MS Adrenomyeloneuropathy, an adult-onset ant of X-linked adrenoleukodystrophy, can particu-larly pose diagnostic difficulties Alexander’s diseaseand cerebral autosomal dominant arteriopathy with subcortical infarcts and leukoencephalopathy(CADASIL) are other rare disorders with featuresimitating MS The recently described vanishingwhite-matter disease only seldom causes confusion
Trang 2vari-with MS With the recent availability of imaging,specific molecular genetic and biochemical tests, thediagnostic dilemma can be easily solved in most ofthese disorders (Kalman and Leist, 2004).
Pregnancy and MS
While family planning may profoundly be influenced
by the level of disability in MS patients, the effect ofpregnancy on the disease has also been a matter ofcontroversy Korn-Lubetzki et al (1984) determined
in a large retrospective study that the frequency ofrelapses decreased during pregnancy, increased inthe postpartum period, and was similar in the preg-nancy year (nine months pregnancy plus threemonths postpartum) to that of out of pregnancy ThePregnancy in Multiple Sclerosis (PRIMS) study was
a large prospective natural history analysis of MS
in pregnant women (Confavreux et al., 1998) Thismulticenter study confirmed the significant decline
of relapse rate during pregnancy, most marked in the third trimester, and the increase of relapse rate
in the first three months postpartum However,
no acceleration of disability was noted during the puerperium, and neither breast feeding nor epiduralanalgesia had negative effects In an extension of thisstudy, patients were followed up to two years post-partum (Vukusic et al., 2004) This second PRIMSstudy added that from the second trimester onwardsand for the following 21 months, the annualizedrelapse rate did not significantly differ from that ofthe prepregnancy year Despite the increased risk inthe first three postpartum months, 72% of womendid not have relapses Increased relapse rate in theprepregnancy year and during pregnancy and ahigher disability status score at pregnancy onset,correlated with the postpartum relapses
3.5 The pathology of MS: A quest for clinical correlation (William F Hickey)
Introduction
Merely a decade ago if one were to delve into thebasic pathology of MS, the picture that would emergewas relatively consistent, but it contained great variability The pathognomonic lesions of MS, calledplaques, were chronic inflammatory foci randomlyaffecting the white matter of the central nervous system that resulted in myelin loss and gliosis (Figs 3.5 and 3.6) The features of the histologicallesions of MS have long been acknowledged to be
highly variable They differed with the age of thelesion and seemed to correlate poorly with the clinicalsyndrome exhibited by the patient (Figs 3.5 and 3.6),except for the fact that if they were situated in theCNS at a specific site, they could be correlated withthe resulting neurological deficit Their histological
Old plaque with myelin loss gliosis and axonal loss (late) B
A
Fig 3.5 The gross photographs (A) of the parietal lobes with totally demyelinated, old MS plaques in the periventricular area on each section Histological sections were prepared from the brain slice in the lower right, and are depicted in B–D B is a Holzer stain to demonstrate gliosis that extends well beyond the area of the MS plaque itself C is a Bodian stain and D is a Bielshowsky stain which identify axons Both demonstrate the near total axonal loss
in the demyelinated zone.
Fig 3.6 The edge of a typical, actively demyelinating MS plaque is shown (hematoxylin and eosin (H&E) stain, X125) The tissue at site “A” is neither inflamed nor demyelinated, and no loss of oligodendroglia has occurred At site “B” loss
of myelin and oligodendroglial cells is nearly complete As indicated by the arrow, the border of the advancing plaque
is hypercellular and contains large numbers of T cells and macrophages.
Trang 3appearance neither helped with prognosis, selection
of therapy nor insight into etiology or pathogenesis
While there had been steady progress in ing the structure of MS plaques using immuno-histochemical and ultrastructural techniques, thefundamental links between the microscopic features
dissect-of the plaque and questions regarding etiology, genesis, and prognosis remained opaque (Hickey,1999; Lassmann, 2005) The various features of thelesions found in MS were reviewed and analyzed
patho-at an internpatho-ational symposium thpatho-at inspected thecomplexity of MS plaques varying from their im-munological constituents and types of damage to the temporal changes in lesional pathology and clinical correlations (Lassmann et al., 1998) It wasobvious that MS was a highly complex, variable, andenigmatic problem
The histopathology of MS has been examined fornearly a century and a half, but progress in under-standing the disease has been slow Pathologistsaccept that while there are certain general features ofthe MS lesion that could be expected based on a lesion’sage, inflammatory activity, and the clinical features
of the illness, a reliable and informative classificationsystem had not evolved if such a system was ever
to prove to be appropriate and useful
Lucchinetti et al (1996) for the first time proposed
a classification schema that apparently permitted
MS cases to be characterized and subdivided basedupon specific immunohistochemical features of thelesions The concept that the pathogenesis of MSmight fall into a set number of specific patterns, eachrepresenting a distinct immunopathological mech-anism, was revolutionary
Pathological subtypes – reality or illusion?
While there have been some minor modifications
in the proposed classification, at this time there arebasically four types of MS lesions that are presumed
to be histologically, immunophenotypically, andpathogenetically distinct (Lucchinetti et al., 2005)
Type I lesions are those characterized by extensiveinfiltration by T cells and macrophages The plaqueshave sharp, distinct edges and the disappearance ofthe various molecular components of myelin seems
to occur simultaneously, not in a selective or ial manner In this type of inflammatory focus someoligodendroglial cells survive the insult and remyeli-nation (partial or complete) may be possible Shadowplaques, areas of incomplete remyelination, can beassociated with type I lesions In many ways type I
sequent-lesions are reminiscent of the pathology found inEAE, a well-established animal model of MS If so,this MS subtype may represent a true autoimmuneattack by T cells against one or more specific myelincomponents
Type II plaques are in many ways similar to theprior type, but are associated with extensive deposi-tion of antibodies and the presence of activated com-plement components, including formation of themembrane attack complex from the final elements
of the complement cascade The lesions have sharpedges and the loss of myelin components occurssimultaneously As before, some oligodendroglialcells are able to survive in the inflammatory foci,thus remyelination can occur and shadow plaquesare found This subtype of MS lesion resembles thepathology of EAE induced by MOG MOG-inducedEAE is distinct in that it requires not only antigen-specific T cells, but also the simultaneous presence
of anti-MOG antibodies Hence, it would seem that
in type II lesions the T cells may be permitting age of antibodies into the CNS, but it is the binding
leak-of antimyelin antibodies and the activation leak-of thecomplement cascade that actually leads to myelindestruction
Type III lesions are distinct from the former two.While there are T cells and macrophages present, the lesions are irregular and the borders ill-defined.Moreover, in this subtype there seems to be a prefer-ential loss of myelin-associated glycoprotein (MAG)over the other molecular components of compactmyelin; in other words, the molecules making upcompact myelin are lost selectively Oligodendroglialcells undergo destruction in what appears to be
an apoptotic fashion, their loss is nearly total andremyelination does not seem possible This MS type
is believed to represent a degeneration of drocytes that starts at their most distal processes.Since at the subcellular level MAG is restricted to the portions of the oligodendroglial processes in theperiaxonal area, it has been suggested that type IIIlesions may represent a “dying-back oligodendro-gliopathy” Such an unusual finding may be parallel
oligoden-to certain features found in hypoxic/ischemic lesions
of the white matter This has led to the hypothesisthat demyelinating foci in some forms of MS mayrepresent hypoxia-like tissue injury (Aboul-Enein
et al., 2005) Indeed, it is possible that some form
of small-vessel vasculitis, possibly one mediated
by activated T cells, may underlie this class of MSdamage (Kornek and Lassmann, 2003; Lassmann
et al., 1998)
Trang 4The type IV lesions in MS are a bit more difficult
to discern Those proposing the classification suggestthat this subtype may represent a distinct disorderaffecting the oligodendroglial cell itself – a so-calledprimary oligodendrogliopathy Histologically, theplaques have sharp edges, are infiltrated by T cellsand macrophages, and the loss of the various myelincomponents appears to occur simultaneously There
is abundant apoptotic death of oligodendroglia in the white matter around the edge of the plaque Yet,the nature of the problem that leads to oligoden-droglial death has not been defined Moreover, thissubtype is rare
There are some problems with this proposed ification system These problems are not necessarilyfatal flaws, nor do they reflect negatively on the proponents who advocate immunophenotypicallycategorizing MS lesions Certainly, few would expectthat an initial classification system based upon a relat-ively small number of cases would be comprehensiveand never need modification or amendment It ismost likely that there will be further refinements ofthe classification based on yet to be identified para-meters Nevertheless, given the proposed classifica-tion system, the current question is whether it should
class-be utilized and broadly applied It is at this point that
we need more data It is coming, but at this writing,not yet available
The aforementioned classification system wasderived from extensive analysis of biopsies from thebrains of patients not previously diagnosed with MSwho presented acutely with a progressive neurolo-gical disorder Some argue that this represents ahighly skewed group of patients, even if the majority(but not all) actually progressed to develop clinicalmultiple sclerosis However, following a furtheranalysis of a broader group of patients, the cat-egorization method appears to be sustained (Pittock
et al., 2005)
Another potential difficulty with the classificationsystem is that it has yet to be replicated and con-firmed by a group not associated with the system’soriginal proponents Access to MS tissue, the scarcity
of MS brain biopsies, the accurate duplication of thereagents and methods used by the original authors,and unfamiliarity with the parameters of analysis
of the tissue employed by the authors of the fication method, are all impediments that must beovercome
classi-One of the major questions concerning the egorization of MS lesions which still remains to beresolved is whether MS lesions are homogeneous
cat-and consistent within an individual patient, or if aspectrum of histopathological types coexists simul-taneously within one person There are reports fromexperienced MS pathologists stating that variousforms of inflammatory lesions do coexist within indi-vidual MS patients (Prineas et al., 2001) Also, studies
of some cases of classic relapsing-remitting MS haveshown lesions that do not neatly fit into the abovecategories (Barnett and Prineas, 2004) Others havereported that there is “notable homogeneity withinindividual patients” (Morales et al., 2006) The answer
is elusive, but should appear in the next few years Atpresent the topic remains a point of much debate.Another final issue with this categorization methodcenters on the extent to which the various histolo-gical types of lesions correlate with specific clinicaltypes of MS It is generally recognized that the clinical course of MS typically falls into a relapsing-remitting pattern, or the secondary progressive type;the primary-progressive form and the so-called
“benign” type are rarer (Lublin, 2005) To date thecorrelation of histopathological type with clinicalsubtype is weak at best (Pittock et al., 2005); how-ever, there are ongoing studies that are specificallydesigned to address this issue of clinical correlation.Can prognosis and therapies be directed by thepathological type? Here there is cause for cautiousoptimism A retrospective study by Keegan et al.(2005) predicted that patients with type II lesions –those characterized by extensive antibody and com-plement deposition – might benefit from therapeuticplasma exchange This is what they found Plasmaexchange did not seem to benefit those with lesiontypes I or III, but individuals with type II lesions experienced moderate to substantial neurologicalimprovement Needless to say, if some correspond-ence between a specific immunohistological patternand a predictable clinical syndrome emerges, thenthe classification of MS based on the lesions’ histo-pathological features will be both broadly acceptedand rapidly applied
Axonal pathology – an unexpected, unifying feature
From the earliest days of the microscopic study of MSlesions, it has been known that axons are damaged
in such lesions But the paper by Trapp et al (1998)still caught those who studied MS unawares Whatwas so amazing was not that axonal damage existed;rather it was the extent to which it was present in
MS lesions Vast numbers of axons were transected
Trang 5in active MS plaques Even in inactive or marginallyactive plaques, the axon damage continued Yet, themost startling observation was that significant axonaldamage was occurring in the normal appearing whitematter, far away from a site of definable inflamma-tion or demyelination (Bjartmar et al., 2001).
While all who study MS agree that axonal damagedoes occur, the puzzle as to whether axonal damagerepresents a primary insult versus a secondary phenomenon is unresolved The work of Trapp andcolleagues strongly suggests that the axonal patho-logy is a unique and primary feature of MS (1998)
Axonopathy may be an early feature in MS lesions(Kornek and Lassmann, 2003) Yet some reports havequestioned this and proposed that axonal damageoccurs in the setting of chronic inflammation andlongstanding disease afflicting the CNS, but is not anecessary or acute phenomenon (Kutzelnigg et al.,2005)
The potential causes of axonal degeneration are manifold Obviously, the presence of a chronicinflammatory infiltrate, activated macrophages andreactive microglial cells, and the elaboration of aspectrum of cytokines and reactive oxygen meta-bolites would create an environment conducive tocell membrane damage (Bjartmar et al., 2000) Thespecific offending entities, however, have not yetbeen specified Alternatively, it has recently beenproposed that mitochondrial dysfunction may be thecause of the axonal damage (Dutta et al., 2006)
Much effort is being expended to dissect this tially critical aspect of MS lesions
poten-The great attention currently being paid to thisseemingly isolated feature of the pathology of MSderives from the fact that many if not all of the fixed neurological deficits found in longstandingcases of MS may result from axonal loss rather than demyelination (Trapp et al., 1999) In cases ofsecondary-progressive MS the constant deteriora-tion of neurological function likewise may be attri-butable to axonal pathology rather than myelin loss
In addition, the relentlessly progressive axonal lossthat seems to occur in MS almost certainly providesthe pathological substrate for the extensive atrophyafflicting all MS patients as they age
Cortical lesions in MS
The existence of focal lesions in the cerebral cortex
of MS patients was a relatively new observation (Bo et al., 2003a) These damaged areas do exhibitgliosis, but are relatively difficult to identify due to
the relative absence of dense myelin in the cortex.Indeed, subpial demyelination can be an extensive, butsubtle, feature in some cases of MS (Bo et al., 2003b).While loss of myelin occurs in cortical lesions, there
is remarkably little inflammatory infiltrate (Bo et al.,2003a) As such, this would suggest that white matter and cortex operate under different ruleswhen it comes to inflammatory demyelination.Perhaps more importantly, this offers the possibilitythat lymphocytes might not be essential in produc-ing damage leading to demyelination, gliosis, andaxonal loss
Even less certain about these cortical and subpiallesions is what they mean clinically Occasional MSpatients exhibit seizures Are such lesions the cause?
Do they contribute to the unusual affect seen in some cases of MS? Can they cause motor or sensoryabnormalities? Again, pathological analysis of theCNS has identified a group of lesions that sporadic-ally do develop in MS, but the clinical phenomenaattributable to such foci are unknown
Summary
In the past decade a system for categorizing thelesions of pathological MS into four discrete subtypeshas been proposed While it is very attractive, somequestion its validity Currently it is not in universaluse because of the uncertainty regarding its ability toprovide any meaningful correlations with etiology,clinical course, prognosis, or therapeutic options At
a deeper level, if the existence of distinct gical patterns of MS plaques is verified and can beemployed by pathologists, do these patterns bespeakdifferent etiologies, different mechanisms, and differ-ent clinical syndromes? Likewise the conundrum
patholo-of whether the CNS lesions are consistently patholo-of thesame type within a given patient throughout thecourse of the disease must be resolved The most elemental and important question regarding MSthat will be answered in the next few years has beenbrought into focus by recent and ongoing patholo-gical analysis of MS tissue Is MS one disease withwidely varying clinical manifestations, or is it actu-ally a number of distinct neuroinflammatory diseaseseach with its own etiology, pathogenetic mechan-ism, and prognosis? It is very possible that the proteandisorder called multiple sclerosis represents a finalcommon pathway for distinct disease entities Withquestions such as this to be resolved the excitementsurrounding the ongoing immunopathological ana-lysis of MS is not likely to abate soon
Trang 63.6 Cerebrospinal fluid (Mark S Freedman)
The cerebrospinal fluid (CSF) or the brain’s “soup”,unlike the blood, is in direct contact with brain cells,hence sampling its contents can give an indication
of what processes may be transpiring in the CNS
In the case of inflammatory conditions such as MS,there are abnormalities that reflect activity arisingfrom within the CNS and help to distinguish themfrom those due to inflammation penetrating the CNS from without An understanding of just whatthe CSF can tell you about inflammatory conditionsthat affect the CNS demands some basic knowledgeabout CSF as well as the limitations of the tests used
to examine it
First it should be pointed out that the blood–brainbarrier (BBB) separating the brain from the vascula-ture is not the same as the blood–CSF barrier (BCB)that comes between the CSF and the blood The BBBtends to be “sealed” by the specialized endothelialtight junctions seen in the CNS, whereas the BCB
is fenestrated acting as a specialized macrofilter
Anything that originates in the blood must crosseither barrier by means of diffusion that is facil-itated either by specialized transporters (e.g pro-teins) or by active transport (e.g glucose) Diffusionacross the BBB is dependent on lipid solubilitywhereas more hydrophilic molecules have an easierpassage through the BCB By measuring the amount
of molecules that are formed outside the CNS, butfound in CSF, it is possible to get some idea of the
“leakiness” of these barriers
Albumin is the simplest molecule measured; formed
in the liver, any amount found in the CSF had tohave traversed the BCB It has long been known thatthe ratio of CSF/serum albumin is a direct measure-ment of BCB permeability (Qalb) which increases withage Using a simple scale, it is possible to estimatewhether permeability is in excess of that expected for
a given age (see Table 3.5)
Conditions that are typically associated with mild
to moderate increase in Qalb include neuropathicprocesses (e.g Guillain–Barré), neuroborelliosis ormeningitis Typically these inflammatory processesare thought to reduce CSF absorption and thereforereduce the natural flow of CSF, which leads to con-centration of albumin within the CSF This reducedCSF flow rate would also lead to intra-CSF accumula-tion of other molecules such as immunoglobulin (Ig).This is the main reason that any measurement ofintrathecal Igs must take into account some meas-ure of BCB leakiness to know if the CSF Ig is simplydue to diffusion in from the blood, or is the directresult of synthesis within the CNS Numerous math-ematical formulas have been devised to account forthis leakiness, and one of the simplest to use is known
as the “Link index” ( Link and Tibbling, 1977):
100% (normal range < 70%)Determining that Ig synthesis had to have arisenwithin the CNS is tantamount to saying that there
is an immune process that is taking place locally.Although this is expected in conditions such as MS,
it is not specific for that disease; rather localized
Ig synthesis is common to any inflammatory CNScondition that leads to humoral immune responses IgG is the commonest Ig to be evaluated, but similar formulas have been used to assess IgA or IgM, thelatter two being of more importance with respect
to infectious causes For instance, in Lyme disease(often considered an important mimic of MS) IgMprevails over IgA or IgG Usually the Qalb is alsomarkedly elevated beyond that expected for age (see Table 3.5) in the case of CNS infectious condi-tions, whereas in MS, it is typically normal Thoughrarely a concern, as dysfunction of the BCB (indicated
by an increase in Qalb) increases, especially due toconditions outside the CNS such as meningitis, for-mulas such as the Link index, which are based on alinear relationship become inaccurate, as the rela-tionship becomes hyperbolic in function and morecomplicated nonlinear formulas are required foraccurately assessing localized Ig synthesis (Reiberand Peter, 2001) The commonest cause for a local-ized increase in Ig is infection However, nonspecificincreases in localized Ig to ubiquitous agents such
as measles, rubella or varicella are common in thepresence of CNS autoimmune-type conditions and
IgG[CSF]/Albumin[CSF]
IgG[serum]/Albumin[serum]
Table 3.5 Increasing values of Qalbwith age.
Age (range) Q alb × 10 −3
Trang 7this so-called “MRZ reaction” (measles-rubella-zoster)typifies the polyspecific nature of Ig activation thattakes place in conditions such as MS (Reiber andPeter, 2001).
Qualitative analysis of CSF Ig is key to the diagnosis
of conditions such as MS It is equally important toinsure that this assessment be performed in a qualifiedlaboratory in a standardized manner (Freedman et al.,2005) There is a clear consensus as to what consti-tutes this analysis (Keir and Thompson, 1990) which
is to perform isoelectric focusing (IEF) of Ig on agarosegels followed by immunoblotting This techniqueseparates the Ig present into either distinct “bands”
suggesting either a specific infection or autoimmuneprocess or into a smear of protein consistent with
a nonspecific increase in Ig It is imperative that comparison be made of CSF Ig directly with serum
Ig, as the presence of bands in CSF that are clearlynot in serum is what constitutes the specificity of the intrathecal response CSF should be applied togels undiluted, whereas serum is usually dilutedempirically 1:400, so as to equate the overall amount
of Ig and minimize overloading in the serum laneswhich can obscure at times the visibility of “bands”
Five patterns of “banding” will emerge using thismethodology (see Fig 3.7) with types II or III beingindicative of intrathecal synthesis of oligcoclonalbanding In most cases, the sensitivity of IEF fordetecting oligoclonal bands in MS is >95% (Paolino
et al., 1996) It should raise an alarm therefore, ifclinical suspicion is high that a patient has MS, butintrathecal synthesis of oligoclonal bands is unde-tected This means that more times than not, ratherthan the test being “falsely negative,” the absence ofoligoclonal bands usually suggests a diagnosis otherthan MS (Zeman et al., 1993)
In considering what the CSF can tell you, it is
important to consider all aspects of CSF analysis:
the cells present (differential or cytology), istry (albumin, glucose, or lactate), as well as the Ig
biochem-These features altogether are used to help guish between causes of systemic inflammationwhich spill over into the CNS, such as vasculitis orchronic infection and intrathecal processes such asthe autoimmune condition MS It is also thereforeimportant to draw simultaneously blood for serumanalysis alongside the CSF, as well as to send it forbiochemical studies, such as glucose Typically 1– 4partially filled tubes of CSF are required and 1–2tubes of blood for full analysis
distin-The first tube can sometimes be contaminatedwith a few red cells from nicking epidural small
vessels during the lumbar puncture The cell countshould be performed no later than two hours afterobtaining the CSF, otherwise changes in cell shapemay hamper the ability to offer a correct and full differential A red blood cell count that is too high(5 – 7 × 109/l ) probably indicates too much of a traumatic tap, rendering other quantitative measure-ments more difficult to interpret If a high number ofred cells are noted in the first tube, then the last CSFtube should also be checked for red cells and if thenumber remains as high as the first tube, then oftenthis is reflective of continued bleeding within thesubarachnoid space such as what might be expected
in a ruptured cerebral aneurysm of arterio–venousmalformation One only needs 1–2 ml of CSF for cellcounts White blood cell counts in the CSF are typic-ally low (normal <5 × 106/l), with any cells presentbeing of lymphocyte origin A single neutrophil seen in a sample free of red cells is cause for concern,possibly indicating either an infection or severe CNSinjury with necrosis Higher than normal whiteblood cell counts have been found in some 34% of MS
CSF S
CSF S CSF S CSF S CSF S
is the pattern seen owing to the presence of a paraprotein (monoclonal IgG component) Courtesy of H Reiber.
Trang 8cases (Tourtellotte, 1970), however, very high counts(>50 × 106/l) are most unusual in MS In some cases,the presence of unusual looking cells should prompt
a full review of cytopathology to exclude the ity of neoplasia or to look for inclusions that mightoccur in certain types of chronic infections such astoxoplasmosis In some cases where a high whitecount is due to lymphocytes, a full tube of 7–10 mlCSF should be drawn and sent for a cytospin and stain-ing with cell markers in order to know for instance
possibil-if the lymphocytes are all B cells, strongly suggesting
a diagnosis of lymphoma, or T cells, more reflective
of either infection or chronic inflammation
For biochemical studies such as glucose, lactate,
or angiotensin-converting enzyme (ACE) 3 – 4 ml ofCSF will usually suffice Low CSF glucose (when com-pared to serum, CSF/serum ratio <0.4) and very hightotal protein content (e.g >1 g/l) is more consistentwith an infectious or neoplastic process Lactate, whereavailable, is a good substitute and has an advant-age over paired CSF–plasma glucose measurements
in that only a single CSF measurement is required(Nelson et al., 1986)
If infectious causes are considered, then a separatesterile tube for Gram stain and microbial or fungalcultures is required Special requests should be made
in cases of chronic meningitis to look for “acid-fastbacillus” and special cultures requested if tuberculosis
is suspected In all cases, if a specific pathogen is suspected, most times specific antigen testing isavailable Regardless, a tube of 3 – 4 ml of CSF is allthat is required for all these analyses
Overall, CSF can be very informative in most cases
of suspected CNS disease A normal CSF in suspectedcases of MS or other possible CNS autoimmune ent-ities is often reassuring and indicates that these dia-gnoses are less likely A typical CSF picture of specificoligoclonal bands in a patient suspected of MS butwho has a MRI that is either normal or shows non-specific lesions and in whom infection has been ruledout would almost certainly turn out to have MS Onthe other hand, the finding of a very high protein, aleaky BCB, or a high cell count in someone who clin-ically is highly suspected of having MS should raiseconcern that a different diagnosis is being missed
A lumbar puncture to obtain CSF along with someserum is a minor procedure with high yields in terms
of reassurance of not missing more treatable tions such as infections, and can help to reinforceclinical certainty of a diagnosis of MS, when clinicalpresentation is somewhat vague or MRI results arenonspecific
condi-3.7 Magnetic resonance imaging characteristics of MS (Jennifer L Cox and Robert Zivadinov)
Introduction
MS is an inflammatory disease of the CNS ized by demyelinating lesions and axonal loss Theimmunopathogenic mechanisms underlying diseaseinitiation and disease course are unknown Currentdiagnostic criteria (McDonald et al., 2001; Polman
character-et al., 2005) suggest MRI is the most sensitive andspecific of the radiological and laboratory tools used
to aid in the diagnosis of MS Although MS could bediagnosed without MRI by waiting for clinical evid-ence of a second attack, it is strongly recommendedthat MRI be used when available to demonstrate dis-semination of lesions in space and time In addition
to its diagnostic usefulness, MRI is routinely used tomonitor the course of MS disease over time
Although conventional MRI scans such as weighted images (WI) and gadolinium (Gd)-enhancedT1-weighted scans have long been used for clinicaldiagnosis and monitoring of MS, they cannot dis-tinguish between inflammation, edema, demyelina-tion, Wallerian degeneration, and axonal loss Inaddition, they do not exhibit a reliable correlationwith clinical measures of disability Some patientshave multiple hyperintense lesions on T2-weightedimages, yet show few clinical symptoms of MS, whileother patients with few hyperintense lesions mayhave a marked clinical presentation The lack of astrong correlation between the presence of lesionsobserved with conventional MRI and clinical symp-toms is often referred to as the “clinical–MRI paradox”(Barkhof, 2002; Zivadinov and Leist, 2005) Further-more, there is increasing evidence that pathologicalchanges in MS can be found in both cortical and subcortical gray-matter structures, yet conventionalMRI scans are not able to detect these gray-matterchanges In recent years, the use of nonconventionalMRI sequences as well as advanced analysis methods
T2-of conventional sequences have allowed the capture
of a more global picture of the range of tissue tions caused by inflammation and neurodegenera-tion Newer, nonconventional metrics of MRI analysisinclude measurement of hypointense lesions on T1-weighted imaging (T1-WI), central nervous sys-tem atrophy, magnetization transfer imaging (MTI),magnetic resonance spectroscopy (MRS), diffusiontensor imaging (DTI), high-field MRI, and functionalMRI (fMRI)
Trang 9altera-When compared to conventional imaging, conventional MRI techniques appear to be bettersurrogate markers for monitoring the destructivepathological processes related to disease activity andclinical progression The nonconventional techniquescan reveal the underlying substrate of intrinsic patho-logy within lesions and normal appearing brain tissue(NABT) that include edema, inflammation, demyelina-tion, axonal loss, and neurodegeneration (Bakshi et al.,2005; Zivadinov and Bakshi, 2004c) Due to theirability to detect the neurodegenerative aspects of
non-MS, including recent evidence for cortical lination (Geurts et al., 2005), these techniques arereceiving increased attention as clinically relevantmarkers of disease progression This section will dis-cuss both conventional and nonconventional MRItechniques and their role in detecting inflammationand neurodegeneration in MS lesions and NABT
demye-Role of conventional MRI in MS
T2-weighted imaging is highly sensitive in detection
of hyperintense lesions in the white matter (WM)and, less commonly, the gray matter (GM) The mosttypical sites for lesions are in the WM: periventri-cular region, corpus callosum, posterior fossa, andcortical regions (Fig 3.8) Several MRI sequences arecapable of identifying T2 hyperintense lesions; thosepreferred most often are conventional spin echo, fastspin echo, and fluid-attenuated inversion recovery(FLAIR) (Zivadinov and Bakshi, 2004c) FLAIR pro-vides improved detection over T2-weighted imaging
in the evaluation of periventricular and cortical/
juxtacortical lesions, as CSF may mask the tion of these plaques on T2-WI (Bakshi et al., 2005;
visualiza-Zivadinov and Bakshi, 2004c) Continuous technicalimprovements in MRI hardware and software overthe last decade have led to the development of moreefficient and sensitive pulse sequences Among them,turbo or fast spin-echo (TSE or FSE) and fast-FLAIRhave already demonstrated their usefulness in awide variety of neurological diseases, including MS(Simon et al., 2006; Zivadinov and Bakshi, 2004c)
FSE has shown greater sensitivity than conventionalspin-echo in detecting areas of T2 prolongation in
MS On the other hand, fast-FLAIR sequences haveemerged as especially helpful in evaluating periven-tricular and cortical/juxtacortical lesions where CSFsignal may mask these plaques on T2-WI (Zivadinovand Bakshi, 2004c) Moreover, double-inversionrecovery (DIR) imaging has recently shown a furtherincrease over FLAIR in the ability to detect cortical
lesions as well as provide better contrast between GMand WM (Geurts et al., 2005) Due to fat suppression,areas of T2 prolongation can also be detected usingshort tau inversion recovery (STIR) sequences and,
in certain scanning platforms, this sequence may besuperior to T2-WI in detecting spinal cord lesions
in MS (Campi et al., 2000) An added advantage
to using STIR when imaging the optic nerves isincreased contrast between lesions and the sur-rounding retrobulbar fat (Moseley et al., 1998).Recently the Consortium of Multiple SclerosisCenters (CMSC) proposed MRI consensus guidelinesfor imaging of the brain and spinal cord in patientswith MS (Simon et al., 2006) Recommended forimaging of the brain were sagittal and axial fast spin-echo fluid-attenuated inversion recovery (fast-FLAIR), axial FSE with proton density (PD) and T2-weighting, and post-Gd-enhanced T1 sequences Anaxial T1-weighted pre-Gd scan and T1-weighted 3Dvolume scan were suggested as optional series toinclude Recommended for imaging of the spinal cordwere sagittal and axial FSE PD-T2 and Gd-enhancedT1 sequences, with a 3D volume scan as optional
Fig 3.8 Axial T2-weighted FLAIR image from a 26-year-old female with relapsing-remitting MS showing periventricular (a) cortical, (b) pericallosal (Dawson’s fingers), (c) hyperintense white-matter lesions (d) Axial T2-weighted FLAIR image from a 25-year-old male with secondary-progressive MS showing hyperintense white matter lesions in the cerebellum and pons.
Trang 10Similar guidelines have also been provided in Europe
by the European Federation of Neurological ScienceTask Force (Filippi et al., 2006)
Despite the sensitivity of T2-WI to reveal diseaseactivity and lesions over time (Paty and Li, 1993),there is only modest correlation between MRI findingsand clinical evolution, except in subjects with veryearly disease (Rudick et al., 2006a; Sailer et al., 1999;
Zivadinov et al., 2001b) Several long-term studieshave examined the correlation of disability pro-gression and the accumulation of T2-lesion burden
One of the longest MRI studies followed patients with clinically isolated syndrome for up to 14 years (Brex et al., 2002) After five years of follow up, datashowed that T2-lesion volume accumulation pre-dicted 25% of the correlation variance in disability,but at 10 years it was down to 16%, and at 14 years,
it explained only about 10 –12% of the variance
Evidence is increasing that diffuse, and particularlycentral, brain atrophy as a characteristic of mid-to-late stage MS may influence this relationship It ispossible that T2-lesion volume may be “artificially”
lowered by the loss of lesions along with normalappearing tissue A decrease in the relationship be-tween T2-lesion volume and disability in advanceddisease stages cautions against the assumptions thatT2-lesion volume progression is a function of diseaseduration alone and that stabilizing T2-lesion volumeindicates a reduction in disease activity (Dwyer et al.,2005; Li et al., 2006)
Despite the previously mentioned limits, severalstrategies for increasing the sensitivity of T2-WIhave become available in the last few years Recentconsensus guidelines recommend a ≤3 mm slicethickness on 2D and ≥1.5 mm on 3D acquisitionsequences for the evaluation of disease burden in MSpatients scanned in clinical routine practice (Simon
et al., 2006) Thinner slices provide increased lesiondetection and higher measurement consistency
Recent consensus guidelines also recommend thatany scanner used in clinical routine practice shouldoperate at a field strength higher than 1.0T With the introduction of 3T MRI systems into clinicalpractice, several questions arise, including the com-parison of 3T versus 1.5T It has been previouslydemonstrated that scanner field strength has a sub-stantial impact on the measured T2 lesion volume(LV), being about 25 – 40% higher with standard 3T magnets than for lower field scanners (Erskine
et al., 2005; Keiper et al., 1998; Sicotte et al., 2003)
Higher-field MRI increases specificity in the tion between detected lesions and clinical disability
correla-Gadolinium enhancement
Gd-enhancement in MS lesions has been connectedwith histopathological findings of the blood–brainbarrier breakdown and active inflammation (Filippi,2000) Gd-enhancing lesions on T1-WI usually cor-respond to areas of high signal intensity on T2-WIand low signal intensity on unenhanced T1-WI, prob-ably due to edema and demyelination associated withthese lesions (Fig 3.9) (Zivadinov and Bakshi, 2004c)
A transient phenomenon in MS, Gd-enhancement isusually detectable for an average of 3 – 6 weeks, andtypically precedes or accompanies the appearance
of a majority of new lesions found on T2-WI in MSpatients Most of the enhancing plaques are not asso-ciated with the presentation of clinical symptomsand do not correlate with clinical status in cross-sectional, and especially longitudinal, studies in themid and long term (Kappos et al., 1999; Zivadinovand Leist, 2005) This discrepancy supports the con-cept that varied factors operate in the occurrence ofrelapses in MS as well as the development of long-term sustained disability Nevertheless, the presence
of continuing enhancement indicates a higher risk
of relapses over the short-to-intermediate term andmay contribute to long-term clinical dysfunction(Filippi, 2000; Zivadinov and Bakshi, 2004c).Several strategies have been proposed to increasethe sensitivity of Gd-enhanced MRI for the detection
of active MS lesions One analysis strategy examinesthe pattern of Gd-enhanced lesions and their relation-ship to lesions found on other MRI sequences Deter-mination of an enhancement pattern may indicatedifferences in the histopathology of MS plaques Con-centric ring-enhancing lesions with central contrastpallor arise in previously damaged areas or in areas
of accelerated local inflammation (Zivadinov and Leist,2005) When compared with homogeneously enhanc-ing plaques, ring-enhancing lesions are larger, have
a shorter duration of enhancement, lower apparentdiffusion coefficient (ADC) and magnetization transferratio (MTR) (Minneboo et al., 2005; Morgen et al.,2001) It has also been shown that ring-enhancinglesions are strong predictors for the development
of persisting hypointense lesions on T1-W1 andbrain atrophy (Bagnato et al., 2003; Minneboo et al.,2005; Zivadinov et al., 2004) Thus, the appearance
of ring-enhancing plaques on Gd-enhanced MRImay not only be characteristic of a more aggressiveform of MS but also predictive of long-term deteri-oration Other strategies that maximize the amount
of information that can be obtained through Gd
Trang 11enhancement include frequent serial monthly ning, scanning the spinal cord as well as the brain,
scan-a delscan-ay of five minutes or more between Gd tion and scanning, using doses of higher contrast(e.g., a double or triple dose instead of a standard 0.1 mmol/kg dose), acquiring thinner tomographicslices, co-registration, reducing the background signal by the application of MTI pulses to T1-WI and,finally, use of high-field strength scanners (Filippi,2000; Zivadinov and Bakshi, 2004c)
injec-Role of nonconventional MRI in MS
Three-dimensional T1-weighted high resolution imaging
The CMSC MRI guidelines suggest the option of collecting high resolution (1 mm × 1 mm × 1.5 mm
voxel size) 3D T1 scans Although high-quality 3D T1scans may take longer to acquire than 2D sequences,they are valuable for many advanced measures ofneurodegeneration in MS, including evaluation ofcross-sectional and longitudinal GM, WM, and CSFvolumes estimates, anatomically defined region ofinterest analyses, and voxel-based morphometry.Mounting evidence supports the idea that brainatrophy is an important biomarker of the diseaseprocess in MS (Fig 3.10) (Bermel and Bakshi, 2006;Miller et al., 2002; Zivadinov and Bakshi, 2004a;Zivadinov and Bakshi, 2004b; Zivadinov and Bakshi,2004d) Several studies emphasize the usefulness ofMRI in assessing CNS atrophy and its relationship tolong-term neurodegeneration and disability progres-sion (Fisher et al., 2002; Zivadinov et al., 2001a)
It has also been established that CNS atrophy is amoderate but significant predictor of neurological
Fig 3.9 Comparison of images from a 25-year-old male with secondary-progressive MS (a, b, c) showing homogeneously enhancing lesions (a) and from a 29-year-old female with relapsing-remitting MS (d, e, f ) showing ring-enhancing lesions (d) (a) and (d): Single dose (0.1 mmol/kg) gadolinium postcontrast axial T1-weighted images (b) and (e): Axial T1-weighted images (precontrast) (c) and (f ): Axial T2-weighted images.
Trang 12impairment (Zivadinov and Bakshi, 2004a; Zivadinovand Bakshi, 2004b; Zivadinov and Bakshi, 2004d).
The association between atrophy and disability isindependent of the effect of conventional MRI lesions
Studies suggest that CNS atrophy begins in patientswith CIS even before the first clinical symptoms,especially in those at high risk for MS The estimatedpercentage change of brain atrophy varies acrossstudies in CIS but is estimated to be 0.8% per year(Zivadinov and Bakshi, 2004a) Natural history andtherapeutic studies of patients treated with placebosuggest that CNS atrophy is common in patients withRR-MS, even in the earliest stages of the disease Theestimated annual rate of whole-brain atrophy variesacross studies but is slightly higher in patients withearly RR-MS (range, −0.7% to 1.33%) than in thosewith advanced RR-MS (range, −0.61% to 1.2%)(Zivadinov and Bakshi, 2004b) Patients with PP-
MS have a slightly higher annual rate of ventricularenlargement (range, −2.4% to +7.7%) than patientswith SP-MS; however, this rate is lower than the rate
in patients with RR-MS (range, +2.1% to +29.8%)
On the other hand, spinal cord atrophy also develops
at a faster rate than brain atrophy in patients withPP-MS (Zivadinov and Bakshi, 2004a; Zivadinovand Bakshi, 2004b)
Although initial reports indicated that brain atrophy in MS was primarily due to decreases in WM(Ge et al., 2000), several more recent reports havenoted diffuse GM atrophy in the brains of patientswith MS (Benedict et al., 2006; Bermel et al., 2003;
Chen et al., 2004; Dalton et al., 2004; Fabiano et al.,2003; Valsasina et al., 2005; Zivadinov et al., 2006)
These findings suggest that the disease process
in MS is not limited to WM and that including GMatrophy in the assessment of patients with MS may
further improve the usefulness of MRI ments Preliminary data from several short- andlong-term studies suggest that GM atrophy develops
measure-at a much faster rmeasure-ate than WM or whole-brain atrophy (Dalton et al., 2004; Valsasina et al., 2005).However, it is not clear whether GM atrophy is aresult of the disease process in MS or is secondary to
WM atrophy
Recently, our research team used an approach based
on regional segmentation called semiautomatic brainregion extraction, or SABRE, to detect predilectionfor brain atrophy development on a region-by-regionbasis (Carone et al., 2006a) The study compared 66
MS patients and 40 normal controls and found thatthe regions most affected in the brain were deepgray-matter structures including the posterior basalganglia and the thalamic regions, as well as the cor-tical regions in the orbital frontal, superior parietal,superior frontal, and medial superior frontal regions.Similarly, reduced thalamic GM volume in patientswith MS was the primary finding in a voxel-basedmorphometry study of CIS patients (Cox et al., 2006).Taken together, these studies suggest that there is aregional specificity for brain atrophy development,prevalent mostly in GM structures, in areas of thebrain that do not usually show WM lesions
Wallerian degeneration and independent neuronaldegeneration are proposed mechanisms for GM atrophy in MS We recently investigated partial correlations between T2 and T1 regional lesion volumes and regional/total GM atrophy in 110 MSpatients (Carone, 2006b) After controlling for total
GM atrophy, partial correlations between regionallesion volume and regional GM atrophy were notsignificant for any of the 26 investigated regions.Results suggest that a distinct generalized disease
Fig 3.10 Axial view of 3D-SPGR image (a) from a 29-year-old female with relapsing-remitting MS An automated, sectional method (structural image evaluation including normalization of atropy, SIENAX) method was applied to the image to generate separate images of gray matter (b), white matter (c), cerebral spinal fluid (d), and segmented image (e).
Trang 13cross-process better accounts for GM atrophy than ally distinct Wallerian degeneration.
region-Furthermore, it is not clear whether GM atrophycontributes to neurological impairment in MS becauseseveral reports have failed to find an associationbetween GM loss and neurological impairment(Dalton et al., 2004; Ge et al., 2000; Sastre-Garriga
et al., 2005) On the other hand, other studies havefound a significant association between GM loss andimpairment, in both RR-MS and PP-MS (Chen et al.,2004; Sanfilipo et al., 2005)
Magnetization transfer imaging, MTI
MTI is an advanced MRI technique that has beenwidely used to evaluate characteristics and evolu-tion of MS lesions and NABT It is based on the inter-actions and exchange between two types of protons:
those that are unbound in a free water pool and thosewhere motion is restricted due to binding with macro-molecules (Filippi and Rocca, 2004) MT contrast isachieved by applying radio frequency (RF) poweronly to the proton magnetization of the macro-molecules Tissue damage in MS is usually reflected
by a reduction in this exchange of protons and thus
a decrease in MTR Decreases in MTR indicate areduced capacity of free water to exchange magnet-ization with the neighboring brain tissue matrix and are not specific to MS pathological substrates
Although MTR decreases are not specific to any of thevarious MS pathological substrates, a relationshiphas been shown between MTR and the percentage
of residual axons and the degree of demyelination(van Buchem et al., 1997) The most compelling evidence in support of this hypothesis comes from apostmortem study that shows a strong correlationbetween MTR values from MS lesions and NABTwith the percentage of residual axons and the degree
of demyelination (Schmierer et al., 2004)
MTI can be used to assess tissue injury in lesions,
in the whole brain and in specific brain structures(Filippi and Rocca, 2004; Sharma, 2006; Zivadinov
et al., 2001a) MTI studies have demonstrated twopossible evolution paths for new MS lesions: (i) insome lesions, a moderate decrease in MTR with subsequent complete recovery of MTR within a few weeks may reflect edema, demyelination, andsubsequent remyelination (Filippi and Rocca, 2004);
(ii) in other lesions, a marked reduction of MTR withonly partial recovery at follow up (Dousset et al.,1998) Different MTI studies have revealed clinic-ally relevant pathological changes in areas of WM
and GM that appear normal on conventional images.Such changes in normal appearing white matter(NAWM) and normal appearing gray matter (NAGM)occur early in the disease process and provide pro-gnostic information pertaining to the risk of MS progression (Filippi et al., 2000; Laule et al., 2003).MTI metrics have been correlated with the degree
of disability (Rovaris et al., 2003) In general, to-strong correlation was found between baselineMTR and subsequent change in the EDSS disabil-ity score These data support the idea that early MTR abnormalities in NABT can predict the clinicalevolution of MS
modest-Magnetic resonance spectroscopy, MRS
MRS offers the potential to investigate tissue structure,metabolism, and function Information about the bio-chemistry of selected brain tissue volumes providespotential surrogate markers for the pathology under-lying MS (Narayana, 2005) MRS imaging allows for the quantitative assessment of inflammation,demyelination, axonal and neuronal injury pro-cesses in MS (Tartaglia and Arnold, 2006) The N-acetylaspartate (NAA) peak in an MR spectrum is
a putative marker of neuronal and axonal integrity,and axonal and neuronal injury can be quantifiedthrough decreases in NAA The choline peak appears
to reflect cell-membrane metabolism (Narayana,2005) An elevated choline peak represents height-ened cell-membrane turnover, as seen in demyelina-tion, remyelination, inflammation, or gliosis MRSfurther provides unique information regarding notonly structural, but also metabolic changes in theCNS Other metabolic peaks of interest in the study
of MS include the glutamate/glutamine peak andmyoinositol peak The glutamate/glutamine peakrepresents a mixture of amino acids and bioaminesused throughout the CNS as excitatory and inhibit-ory neurotransmitters (Srinivasan et al., 2005) Themyoinositol peak represents a sugar-like moleculethought to be a marker of glial proliferation and nowrecognized for its importance in osmotic regulation
of brain tissue volume (Narayanan et al., 2006).Recent MRS studies have shown that neurodegen-erative changes may be detected in cortical lesionsand deep GM tissue (Geurts et al., 2006; Inglese et al.,2004) Correlations between disability and decreasedNAA–creatine (NAA–Cr) ratio were found in severalstudies, suggesting that MRS measures of brainmetabolites are better predictors of clinical disabilitythan conventional MRI