Document Sample
RESEARCH REVIEWS 14: 221 – 228 (2008)

                 AFFECTIVE DISORDER
                             Elizabeth M. Wells,1 Karin S. Walsh,2 Zarir P. Khademian,3
                                      Robert F. Keating,4 and Roger J. Packer1*
                                     Division of Neurology, Brain Tumor Institute, Children’s National Medical Center,
                      The George Washington University School of Medicine and Health Sciences, Washington DC
                            Division of Neuropsychology, Brain Tumor Institute, Children’s National Medical Center,
                      The George Washington University School of Medicine and Health Sciences, Washington DC
                             Division of Neuroradiology, Brain Tumor Institute, Children’s National Medical Center,
                      The George Washington University School of Medicine and Health Sciences, Washington DC
                             Division of Neurosurgery, Brain Tumor Institute, Children’s National Medical Center,
                      The George Washington University School of Medicine and Health Sciences, Washington DC

        The postoperative cerebellar mutism syndrome (CMS), consisting            (PFS), CMS is characterized by severely diminished or absent
of diminished speech output, hypotonia, ataxia, and emotional lability,           speech output as well as other neurological, cognitive, and be-
occurs after surgery in up to 25% of patients with medulloblastoma and
occasionally after removal of other posterior fossa tumors. Although the
                                                                                  havioral impairments. In addition to mutism, the most com-
mutism is transient, speech rarely normalizes and the syndrome is associ-         monly observed findings are ataxia, hypotonia, and emotional
ated with long-term adverse neurological, cognitive, and psychological            lability. First described in 1979, CMS has been reported in
sequelae. The clinical, neuroradiographic, and neuropsychological find-           over 300 cases to date [Gelabert-Gonzalez and Fernandez-
ings associated with CMS as well as possible mechanisms of injury are             Villa, 2001; Robertson et al., 2006]. Onset is typically 1–2
reviewed. Theories about the pathophysiology of CMS have evolved
along with our understanding of the cerebellum as an important struc-             days postresection. Although the syndrome was previously
ture in the distributive neurocircuitry underlying complex speech, cogni-         considered transient, recent large scale studies reveal persistent
tion, and behavior. CMS shares many similarities with the cerebellar cog-         features and poor long-term outcomes. To date, no demo-
nitive affective syndrome, more commonly described in adults and con-             graphic, clinical, or radiographic features consistently predict
sisting of disturbances of executive function, visuospatial skills, nonmotor
language, and affect regulation. Future directions include more thorough
                                                                                  which patients will develop the syndrome. This manuscript
neuropsychological characterization, functional and diffusion tensor              will review the clinical, neuroradiographic, and neuropsycho-
imaging studies, and investigations into the underlying differences that          logical findings associated with the syndrome, both at the time
may make some patients more vulnerable to CMS.          ' 2008 Wiley-Liss, Inc.   of diagnosis and at long-term follow-up. Theories concerning
Dev Disabil Res Rev 2008;14:221–228.
                                                                                  the pathophysiology and proposed mechanisms of injury, and
                                                                                  the implications for cerebellar cognitive function will be dis-
Key Words: cerebellar mutism; posterior fossa; cerebellar cognitive               cussed, as will the relationship to the cerebellar cognitive
affective syndrome; medulloblastoma; cerebellum                                   affective syndrome (CCAS), a chronic condition more com-
                                                                                  monly described in adults with cerebellar damage.

                                                                                   *Correspondence to: Roger J. Packer, Neuroscience and Behavioral Medicine,

        he postoperative cerebellar mutism syndrome (CMS) is                       Department of Neurology, Brain Tumor Institute, Children’s National Medical Cen-
        a debilitating condition that develops in a subset                         ter, 111 Michigan Avenue, NW, WA DC 20010. E-mail:
        patients who have undergone resection of posterior                         Received 3 September 2008; Accepted 3 September 2008
                                                                                   Published online in Wiley InterScience (
fossa tumors, including up to 25% of patients with medullo-                        DOI: 10.1002/ddrr.25
blastoma. Also referred to as the posterior fossa syndrome
' 2008 Wiley -Liss, Inc.
INCIDENCE                                    and vision loss without dysfunction of          92% with ataxia, 66% with speech and
       In case reports published in the      the third cranial nerve has also been           language dysfunction, and 59% with
1980s and 1990s, CMS was usually             reported [Dailey et al., 1995; Liu et al.,      global intellectual impairment. Of the
described as rare. Single institution ret-   1998; Siffert et al., 2000; Ozgur et al.,       52 patients with moderate CMS, 78%
rospective studies of postoperative CMS      2006]. Personality and emotional distur-        had ataxia, 25% had speech and lan-
in children with posterior fossa tumors      bances includes irritability, disinhibition,    guage dysfunction, and 17% had global
have found an incidence ranging from 2       inattention, and lability of affect with        intellectual impairment. Thus, impair-
to 40% [Van Calenbergh et al., 1995;         poor cognitive and behavioral modula-           ment in these domains was common
Pollack, 1997; Robertson et al., 2006;       tion [Levisohn et al., 2000; Robertson          and was also directly related to the se-
De Smet et al., 2007]. A recent pro-         et al., 2006]. Cognitive impairment also        verity of CMS.
spective study by the Children’s Oncol-      occurs and will be discussed below.
ogy Group (COG) of 450 children with                Until recently, CMS was believed         RISK FACTORS
medulloblastoma found CMS following          to be a transient phenomenon. Most                     Investigators have tried to identify
tumor resection in 24% [Robertson            early case studies describe the mutism as       risk factors in the development of
et al., 2006]. Most cases of CMS occur       relatively brief, lasting days to months        CMS. Tumor type has been identified
following resection of posterior fossa       on average. In a 2001 review, mutism            as a risk factor, with higher rates of
tumors; however, it has also been            was described as transient in all 134           CMS observed in medulloblastoma
described in other settings, both surgical   cases examined [Gelabert-Gonzalez and           compared to other pediatric posterior
and nonsurgical. Such cases include an       Fernandez-Villa, 2001]. However, the            fossa tumors. Vermal location has often
acute subdural posterior fossa hematoma      report did not characterize long-term           been associated with the syndrome [Van
requiring excisions [Fujisawa et al.,        speech and language outcomes or other           Calenbergh et al., 1995; Grill et al.,
2005], a cystic hemangioblastoma [Akil       neurological and cognitive features of          2004], and in a 2001 review, 89% of
et al., 2006], an arteriovenous malfor-      CMS. More recent studies have demon-            134 cases had the vermis as the tumor
mation of the vermis [Al-Anazi et al.,       strated that while some aspects of the          site [Gelabert-Gonzalez and Fernandez-
2001], posterior circulation infarction      syndrome may be fleeting, persistent            Villa, 2001]. There has been no associa-
[Nandagopal       and     Krishnamoorthy,                                                    tion between age and gender at the
2004], and acute cerebellitis [Papavasi-                                                     time of diagnosis and the development
liou et al., 2004]. The syndrome has                                                         of CMS [Grill et al., 2004; Robertson
been described in both pediatric and                                                         et al., 2006; Turgut, 2008]. Although
adult literature, but it is more common
                                              More recent studies have                       previous case studies indicated tumor
in children [Ildan et al., 2002].             demonstrated that while                        size as a risk factor [Gelabert-Gonzalez
                                                                                             and Fernandez-Villa, 2001], larger stud-
CLINICAL FEATURES                               some aspects of the                          ies have not supported this theory
      In patients with CMS, loss of              syndrome may be                             [Robertson et al., 2006, Wells et al.,
speech occurs without corresponding                                                          in press]. Damage to the dentatothalao-
long tract signs, supranuclear or cranial       fleeting, persistent                         cortical outflow tracts from the cerebel-
nerve palsies. Most reports of postoper-      impairment is common.                          lar nuclei has also been identified as a
ative CMS describe a latent phase of                                                         possible risk factor. The COG study
1–5 days prior to the onset of symp-                                                         found that brainstem involvement was
toms, although in clinical practice early                                                    more likely to occur in patients with
detection following surgery is difficult                                                     CMS compared to patients without
and onset is earlier in many patients,       impairment is common, with complete             CMS after resection of a posterior fossa
but poorly documented. Resolution of         recovery of speech and language being           tumor [Robertson et al., 2006]. Review
the muteness is often followed by a          infrequent [Siffert et al., 2000; Steinbok      of causation from a series of patients
period of dysarthria [Van Calenbergh         et al., 2003; Huber et al., 2006; Rob-          seen at the Children’s National Medical
et al., 1995; Vadeinse and Hornyak,          ertson et al., 2006]. Motor speech defi-        Center revealed a trend, but no statisti-
1997; Riva and Giorgi, 2000]. Dysarth-       cits following the period of mutism             cally significant associations between
ric speech in patients with cerebellar       have been identified, and the more              mutism and brainstem involvement
lesions has been described as loud, mo-      chronic sequelae have been termed cer-          [Wells et al., in press].
notonous, scanning, slurring and with a      ebellar mutism and subsequent dysarth-
slow rhythm and fluctuation of pitch,        ria [De Smet et al., 2007]. Some of the         NEUROANATOMIC SUBSTRATE
consistent with an ataxic dysarthria [Van    associated behavioral disturbances have               CMS has been associated with
Dongen et al., 1994; Ozimek et al.,          also been described as transient [Levi-         lesions of the cerebellar vermis, lateral
2004].                                       sohn et al., 2000].                             hemispheres, dentate nuclei, and the
      Patients with CMS/PFS are pres-               More recent research suggests that       connections between these structures
ent with several neurological, cognitive,    the neurological and cognitive impair-          and the cerebral cortex. Much attention
and behavioral features in addition to       ments in CMS often persist. A COG               has been paid to the cerebellar vermis,
the mutism and subsequent dysarthria.        prospective study evaluated the neuro-          because of the high frequency of mid-
The most common neurological find-           logical status of patients 1 year postdiag-     line tumors in patients with CMS and
ings are ataxia and hypotonia. Recovery      nosis based on the presence and severity        the absence of mutism in patients with
of gait and coordination is reported to      of ataxia, language difficulties, and other     large cerebellar hemispheric tumors that
take longer than speech [Siffert et al.,     cognitive deficits [Robertson et al.,           have been resected without directly
2000; Steinbok et al., 2003]. Oropha-        2006]. Of the 46 patients who had               involving the vermis [Grill et al., 2004].
ryngeal apraxia, hypokinesis, horizontal     postoperative CMS initially rated severe,       However, resection of a vermian tumor
gaze paralysis, persistent eyelid closure,   residual deficits were common including         does not necessarily lead to mutism. In
222                                                        Dev Disabil Res Rev         Cerebellar Mutism Syndrome         WELLS ET AL.
addition, CMS can result from damage               Functional imaging of CMS is           trauma or vasospasm causing delayed
to adjacent structures [Pollack et al.,      limited. A positron emission tomogra-        injury to the cerebellar vermis, brain-
1995]. The dentate nuclei are known to       phy (PET) study showed that patients         stem, dentate nuclei, or their projec-
be involved in initiating voluntary          with mutism may have decreased cere-         tions.
movements, and damage to these para-         bral blood flow in the thalami, medial               Postsurgical edema from intra-
vermian structures is frequently seen in     frontal lobes, temporal lobes, and cere-     operative manipulation has been sug-
patients with CMS [Ozgur et al., 2006].      bellar vermis [Sagiuchi et al., 2001].       gested to be the principle mediator of
However, CMS does occur in the ab-           However, in Pollack’s study of 12 chil-      cerebellar mutism. Some studies have
sence of edema or infarction in the          dren experiencing CMS, PET was unre-         shown increased edema in the cerebel-
dentate nuclei [Dailey et al., 1995], sug-   markable [Pollack et al., 1995]. SPECT       lum in patients with CMS [Dietze and
gesting additional pathways. Other           has demonstrated hypoperfusion in the        Mickle, 1990; Ersahin et al., 1996]. The
structures in the afferent and efferent      left cerebellar hemisphere in a patient      delayed onset of the speech impairment
pathways between the dentate nuclei          with CMS, and the radiographic abnor-        may reflect the time interval needed for
and the premotor and supplementary           mality resolved along with the mutism        edema to spread to the critical struc-
motor cortices have also been impli-         [Ersahin et al., 1996].                      tures. Focal postoperative edema of the
cated in the syndrome [Pollack et al.,             In a study currently being com-        cerebellar peduncles and brain stem has
1995; Koh et al., 1997; Ozgur et al.,        pleted at the Children’s National Medical    been found to be a distinguishing fea-
2006].                                       Center, a variety of neuroradiographic       ture of CMS compared to other postop-
                                                                                          erative cases in one study [Pollack et al.,
                                                                                          1995], but this finding was not repli-
NEURORADIOLOGIC STUDIES                                                                   cated [Robertson et al., 2006]. Another
       Investigators have tried to deter-      Neuroimaging studies                       theory is that vasospasm and subsequent
mine whether neuroimaging may be              support the theory that                     postoperative ischemia may account for
used to predict the occurrence or out-                                                    the delayed onset and possible transient
come of CMS. Although case reports                  the underlying                        nature of mutism [Turgut, 1998; Al-
reveal postoperative damage in several                                                    Anazi et al., 2001], however this study
neuroanatomic areas, no abnormality
                                               neuoranatomical locus                      has not been supported by radiographic
has been consistently found in CMS. A                 may be the                          findings. Perhaps, studies employing dif-
few systematic neuroimaging studies                                                       fusion tensor imaging will help clarify
have compared patients with and with-          dentatothalamocortical                     the type of cellular damage in CMS.
out mutism, following posterior fossa          outflow tracts from the                            Hydrocephalus has also been
tumor removal. Neither the size of the                                                    hypothesized to be related to CMS.
tumor, presence of hydrocephalus, nor        cerebellar nuclei through                    Upon noting that resolution of hydro-
the length of vermian incision was asso-        the brainstem. They                       cephalus coincided with improvement
ciated with the development of mutism                                                     of mutism in many cases, Turgut postu-
[Robertson et al., 2006; Wells et al.,             suggest that the                       lated that CMS is caused by delayed
in press].
       The COG prospective study
                                             occurrence of CMS is not                     transient ischemia secondary to hydro-
                                                                                          cephalus and sensitivity of Purkinje cells
found that patients with CMS had more           attributed to just one                    [Turgut, 2007]. However, other studies
brainstem invasion compared to patients        neuroanatomic region,                      have failed to identify hydrocephalus as
without CMS. There was also a nega-                                                       a specific risk factor and recent reports of
tive correlation between CMS and cer-         but to any disruption of                    long-term sequelae contradict the tran-
ebellar hemispheric tumors [Robertson                                                     sient aspect of the argument [Robertson
et al., 2006]. Siffert et al. [2000] com-
                                                this neural circuitry.                    et al., 2006].
pared CT and MRI scans in eight chil-                                                             It is not known if or how surgical
dren with CMS to eight unaffected                                                         intervention during excision of cerebel-
children and found no distinguishing                                                      lar tumors results in the damage that
features. Pollack et al. performed a         features have been analyzed. This study      causes CMS. Some investigators believe
blinded comparison of CT or MRI              suggests that CMS is associated with last-   that more aggressive surgical technique
scans of 12 patients with mutism to 24       ing postoperative damage to the cerebel-     may determine who develops CMS. Di
cases of vermian tumors without              lum and brainstem, and this damage is not    Cataldo et al. [2001] pointed out that
mutism and found that bilateral edema        identified on immediate postoperative        CMS had not been reported prior to
within the cerebellar peduncles postop-      MRI [Wells et al., in press]. The changes    1985, and speculated that improvements
eratively was the only factor that was       include atrophy in all or parts of the       in imaging and surgical techniques have
significantly associated with the mutism     cerebellar vermis and/or cerebellar hemi-    made neurosurgeons more willing to
syndrome [Pollack et al., 1995]. These       spheres.                                     attempt interventions previously judged
neuroimaging studies support the                                                          impossible. CMS has been shown to
theory that the underlying neuorana-         CAUSE/MECHANISM OF                           occur more often in patients with radi-
tomical locus may be the dentatothala-       INJURY                                       cal resection than in patients with resid-
mocortical outflow tracts from the cere-           The mechanism of injury in post-       ual tumors [Van Calenbergh et al.,
bellar nuclei through the brainstem.         operative CMS continues to be a matter       1995]. The COG prospective study an-
They suggest that the occurrence of          of debate. Given the absence of symp-        alyzed the occurrence of CMS associ-
CMS is not attributed to just one neu-       toms preoperatively, damage from the         ated with the performance of surgery
roanatomic region, but to any disrup-        tumor itself is unlikely. Possible mecha-    by general and pediatric neurosurgeons,
tion of this neural circuitry.               nisms include direct injury from surgical    and found no difference.
Dev Disabil Res Rev       Cerebellar Mutism Syndrome        WELLS ET AL.                                                       223
       Initially, inferior vermian incisions   NEUROCOGNITIVE FEATURES                          old with medulloblastoma and associ-
made during tumor excision were con-                   As the survival rate for children        ated hydrocephalus. One day postsurgi-
sidered possible causes for the develop-       with brain tumors has continued to rise,         cal resection, the patient exhibited cere-
ment of mutism. A single institution           a focus on longer-term effects of the            bellar dysmetria, dysdiadokinesia, and
case series found a correlation between        disease and associated treatment has             mutism. Although the motor symptoms
CMS and this surgical approach [Dailey         emerged. The known neurocognitive                continued to improve over the next
et al., 1995]. However, changes in sur-        effects and neuropsychological evalua-           several weeks, the mutism remained.
gical technique that avoid splitting of        tions have become increasingly consid-           Serendipitously, the patient was exposed
the vermis have not demonstrated the           ered and incorporated into the standard          to familiar and favorite music and began
prevention of CMS [Siffert et al.,             of care for survivors of cancers.                singing without prompts, but remained
2000]. Furthermore, superior and infe-                 Only relatively recently have there      mute without the music. However, the
rior vermian incisions have both been          been any studies specifically designed to        patient’s speech recovered quickly there-
associated with postoperative mutism           assess the neurocognitive outcomes of            after. Other symptomatology during
and it may occur even without damage           children who have demonstrated post-             that period included decreased initia-
to median and paramedian structures            surgical complications such as CMS, and          tion, poor regulatory control and atten-
[Ersahin et al., 2002]. Moreover, the          the majority have been retrospective             tion, impaired language comprehension,
majority of children with midline poste-       chart reviews. Nonetheless, several neu-         and emotional apathy and irritability.
rior fossa tumors undergo radical resec-       rocognitive domains have been docu-                     In another study, children with
tion through the vermis, and only 25%          mented to be disrupted in children who           resected medulloblastoma (who rou-
or fewer develop CMS. The use of sur-          have undergone posterior fossa tumor             tinely receive radiotherapy and chemo-
gical retractors has not been studied,         resection with or without subsequent             therapy after surgery) have also been
although it has been raised as another         CMS. The existing literature in children         identified as having poorer long-term
possible cause [Ozgur et al., 2006]. It is     has focused on posterior fossa astrocy-          cognitive outcome compared to chil-
possible that subtle variations in surgical    toma, medulloblastoma, and more rarely           dren with astrocytoma (who are usually
approach may alter the risk of CMS,            ependymoma. Some research has com-               treated with surgery, alone), specifically
however, a means of preventing CMS             pared benign and malignant tumors as             with sustained attention, long-term vis-
has not been found.                            to control for the effects of radiation          ual memory, and various executive
       Some investigators have argued          and chemotherapy on the longer-term              functions including set-shifting or cog-
that the neurocognitive problems               neurocognitive effects.                          nitive flexibility and working memory
described in patients with CMS may be                  In a retrospective study, Riva and       [Ronning et al., 2005]. Both groups
attributed, at least partly, to the effects    Giorgi [2000] compared consecutive               demonstrated long-term impairments.
of radiation or chemotherapy [Review           patients with cerebellar hemispheric             Of note, some of these differences may
in Levisohn et al., 2000]. Radiation ne-       astrocytoma to those with cerebellar             be attributable to radiation effects, as
crosis is associated with deficits in gen-     vermis medulloblastoma within 6 weeks            the medulloblastoma group had been
eral intelligence, processing speed,           of surgery and prior to any additional           treated prior to the assessment period.
attention and executive functions,             treatment (e.g., radiation or chemother-         In this particular study, there was a sig-
visuomotor control, as well as academic        apy). The patients with nonmidline               nificant age effect only in the medullo-
achievement, with significant sequelae         tumors (e.g., astrocytomas) demon-               blastoma group such that younger age
observed most prominently in younger           strated a striking lateralizing effect           at diagnosis resulted in poorer outcome.
children [Packer et al., 1989; Palmer          related to verbal (global receptive and                 Aarsen and colleagues studied a
et al., 2007; Mabbott et al., 2008].           expressive language deficits), nonverbal         group of children with cerebellar pilo-
Methotrexate chemotherapy, most nota-          skills (visuospatial fluency), as well as se-    cytic astrocytoma without radiation or
bly intrathecal administration, has also       quential memory (verbal and visuospa-            chemotherapy treatment in a retrospec-
been associated with later neurological        tial), which has been hypothesized, but          tive study [2004]. The group exhibited
and neurocognitive impairments. How-           not consistently demonstrated in the lit-        long-term impairments in divided and
ever, since both radiation and chemo-          erature. However, both right and left            selective attention, processing speed,
therapy are administered in standard           cerebellar tumor groups exhibited nota-          visuomotor skills, visual memory, and
doses and schedules and CMS occurs in          ble executive dysfunction, specifically          higher order initiation and set-shifting
a quarter or less of treated patients, it is   related to verbal initiation, planning,          or flexibility. A lateralizing effect was
unlikely that these treatments are the         and set-shifting. Deficits in processing         also reported, such that patients with
root cause. Moreover, the onset of             speed were also exhibited.                       right cerebellar hemisphere tumor
symptoms is typically days after surgery,              Six of the eleven medulloblastoma        resection exhibited language and verbal
before radiation or chemotherapy has           patients exhibited CMS, and those with           memory impairments, while those with
been initiated. A recent study showed          residual language deficits also demon-           left hemisphere resection resulted in
that patients who received cranial radia-      strated additional executive function def-       impaired visuospatial and nonverbal
tion for posterior fossa tumors had            icits including poor set-shifting and plan-      memory functions. Overall, those with
demonstrated impairments, but those            ning. The remainder of the 11 patients           left hemisphere tumors fared worse
with postsurgical complications in addi-       were classified as having ‘‘behavioral dis-      globally. With regard to developmental
tion to CRT had the poorest outcomes           turbance’’ without mutism following sur-         effects, Aarsen reported a significant age
[Mabbott et al., 2008]. It has been sug-       gery, however, they also exhibited prob-         effect for children diagnosed between
gested that children who develop post-         lems with expressive language and exec-          the ages of seven and nine and a half
surgical mutism may be at greater risk         utive functioning, specifically related to       compared to those diagnosed prior to
for long-term sequelae secondary to            verbal initiation and set-shifting.              seven or later than ten.
damage by radiation and chemotherapy                   In a unique case study of CMS,                  Several other studies have demon-
[Glauser and Packer, 1991].                    Ozgur et al. [2006] described a 5-year-          strated similar late effects in patients
224                                                          Dev Disabil Res Rev          Cerebellar Mutism Syndrome         WELLS ET AL.
with CMS, including diminished proc-            lesions of the vermis were highly associ-            Criteria for CCAS in children are
essing speed, poor verbal initiation and        ated with dysregulation of affect, con-       not yet defined. When Levisohn et al.
other language deficits, impaired atten-        sistent with the conceptualization of the     described CCAS in their pediatric pop-
tion, and executive functions (e.g., set-       vermis as the ‘‘limbic system’’ of the cer-   ulation, they included patients with
shifting, novel problem solving), as well       ebellum [Schmahmann, 1991]. While             some, but not all of the characteristic
as memory impairment [Vadeinse and              this was a retrospective study, it suggests   set of behaviors [Levisohn et al., 2000].
Hornyak, 1997; Aarsen et al., 1995].            that the manifestation of CCAS is evi-        Reports of the neuropsychological and
       It is difficult to ignore the similar-   dent in children and, in the acute setting    behavioral consequences of CMS dem-
ities between CMS and the CCAS.                 after posterior fossa surgery, they overlap   onstrate variations in the affected cogni-
CCAS consists of a set of nonmotor              with those of CMS. It underscores the         tive domains. The extent to which
deficits that are found in patients suffer-     role of the cerebellum in the distributed     there is overlap between patients with
ing from different cerebellar abnormal-         neural circuitry subserving higher-order      CMS and features of CCAS is not yet
ities and lesions. Schmahmann and               cognitive and behavioral functions.           known. Mutism appears to be a key
Sherman [1998] defined the syndrome                    The      key     components       of   feature of the initial presentation of
after they performed neurological               CCAS—disturbances of executive func-          many children with CCAS, particularly
examinations, bedside mental status             tion, visuospatial skills, nonmotor lan-      if the children have damage to the cere-
examinations, neuropsychological stud-                                                        bellar vermis. More neuropsychological
ies, and anatomical neuroimaging on 20                                                        and functional neuroimaging studies of
patients with diseases confined to the                                                        well characterized groups of children
cerebellum. Patients with CCAS dem-              Mutism appears to be a                       and adults with CCAS and CMS may
onstrate impairment of executive func-          key feature of the initial                    shed light on the nature of these syn-
tions such as planning, set-shifting,                                                         dromes and the extent to which their
verbal fluency, abstract reasoning, and            presentation of many                       clinical overlap is anatomically based.
working memory; difficulties with visual-                                                     Executive functional impairment, a cen-
spatial organization and memory; per-
                                                   children with CCAS,                        tral feature of CCAS in adults, has been
sonality change with blunting of affect              particularly if the                      poorly characterized in children with
or disinhibited and inappropriate behav-                                                      CMS and needs further study.
ior; and language deficits including             children have damage to
agrammatism and dysprosodia. CCAS is               the cerebellar vermis.                     WHAT CMS TEACHES US
attributed to disruption of the cerebellar                                                    ABOUT CEREBELLAR
modulation of neural circuits that link         More neuropsychological                       FUNCTION
prefrontal, posterior parietal, superior               and functional                                In recent years, the role of the
temporal and limbic cortices with the                                                         cerebellum in higher cognitive functions
cerebellum. These impairments and the            neuroimaging studies of                      has become increasingly clear. There is a
theory behind them may apply to
patients with the CMS as well.
                                                well characterized groups                     complex circuitry between the cerebel-
                                                                                              lum and cerebral cortex, and diaschisis
       Levisohn et al. [2000] evaluated            of children and adults                     secondary to disruption of this circuitry
neuropsychological data in 19 children            with CCAS and CMS                           has been proposed to be the likely etiol-
referred for neuropsychological assess-                                                       ogy of neurocognitive, psychological,
ment following resection of cerebellar             may shed light on the                      and behavioral alterations observed in
tumors ranging in age from 3 to 14                                                            CMS. Once considered solely responsi-
years. The patients were included if the
                                                nature of these syndromes                     ble for coordination of purposeful motor
neurocognitive assessment was com-               and the extent to which                      movement, the contribution of the cere-
pleted prior to irradiation therapy, and                                                      bellum to higher-order cognitive func-
there was no history of premorbid cog-            their clinical overlap is                   tion such as language, cognitive and be-
nitive dysfunction. Although not a                  anatomically based.                       havioral regulatory control, and execu-
selection criteria, none of the patients                                                      tive functions (including initiation,
received methotrexate chemotherapy.                                                           novel problem-solving, working mem-
Of the 19 patients included in this study,                                                    ory, and cognitive flexibility) have been
the majority had medulloblastoma                guage, and affect regulation—have all         demonstrated [Review in Riva and
(58%), while astrocytoma (37%) and              been described in patients with CMS.          Giorgi, 2000]. This understanding has
ependymoma (5%) constituted the                 The major distinction between CCAS            evolved over the past 20 years and has
remaining patients. Although the disease        and CMS appear to relate to the chron-        been supported by case reports and
etiology of the children is unclear, 56%        icity of the symptoms, with the latter        increasingly more sophisticated neuroi-
of the 19 children had extensive ver-           seemingly more transient. However,            maging studies [Leiner et al., 1986,
mian damage and exhibited CMS as                patients who have been described as           Middleton and Strick, 1994, Steinlin,
well as additional neuropsychological           having CCAS have undergone thor-              2007, Timmann and Daum, 2007].
impairments consistent with those               ough neuropsychological evaluation,           Functional MRI and PET studies have
reported in CCAS in adults. Specifically,       allowing for the identification of more       shown activation of the cerebellum dur-
the authors noted impairments in global         specific neurocognitive deficits, whereas     ing nonmotor tasks, including executive
intellectual functioning, aspects of exec-      this has rarely been studied in children      [Bellebaum and Daum, 2007], expres-
utive function, visual-spatial function,        with CMS, as much attention has been          sive language [Petersen et al., 1988], and
expressive language, verbal memory, and         paid to the transient loss of speech          verbal memory tasks [Andreasen et al.,
modulation of affect during the first 2         rather than the potential cognitive           1995]. The role of the cerebellum in
years after surgery. They found that            sequelae [Sadeh and Cohen, 2001].             emotional behavior has been demon-
Dev Disabil Res Rev         Cerebellar Mutism Syndrome          WELLS ET AL.                                                     225
strated in primates with cerebellar           cerebellar-pons-cerebrum circuit. Many         to the higher incidence of damage to
lesions. Other evidence that the cerebel-     cases of cerebellar abnormalities involve      the cerebellar vermis and nearby nuclei
lum plays a role in higher-order behav-       not only cerebellar damage, but also           in children, as posterior fossa tumors
iors comes from imaging studies of chil-      cerebral cortical structural or functional     including medulloblastoma are much
dren with neuropsychiatric and genetic        abnormalities. However, the develop-           more common in children than adults.
disorders such as Attention Defict/           ment of cerebellar tumors usually occurs       Mutism may also be related to develop-
Hyperactivity Disorder (ADHD), au-            without cerebral damage, either preor          mental factors including the absence of
tism, developmental dyslexia, Fragile X,      postoperatively. MRI studies have              well-myelinated pathways for initiation
Down syndrome, and schizophrenia.             shown very few cases of damage to the          of speech and other behaviors.
[Chang et al., 1998; Mostofsky et al.,        cerebrum in patients with postoperative        Although lack of myelination may make
1998; Giedd et al., 2001; Nicolson            CMS. Thus, CMS enables us to exam-             them prone to surgical injury from
et al., 2001; Pinter et al., 2001; Eckert     ine outcomes in patients without cere-         either direct insult or retraction, it may
et al., 2003; Cornish et al., 2005; Stood-    bral damage and further supports the           also make them more capable in recov-
ley et al., 2006; Zang et al., 2007].         importance of the posterior fossa in           ery and redirection [Ozgur et al., 2006].
        Initially, case reports of CMS        both motor coordination and higher-
focused on the motor speech aspect. At        order cognitive function.                      CONCLUSION
a time when neuroscience focused on                  Levisohn et al. [2000] found that               Tumors of the posterior fossa
functional domains and the cerebellum         patients with CMS including affective          account for half of all brain tumors in
was primarily understood to coordinate        changes also demonstrated cognitive            children [Packer et al., 2008]. As the
movements, researchers proposed that a        impairment, but patients with cognitive        prognosis for survival is improving as a
loss of coordination of orofacial move-       changes did not necessarily show CMS           result of advances in treatment, clini-
ments is mainly responsible for the de-       and affect disturbance. This finding is        cians and researchers are paying greater
velopment of mutism. This theory was          consistent with the hypothesis that            attention to the long-term neurocogni-
supported by reports that the time            affect regulation is principally a function    tive and psychological sequelae associ-
course of development and recovery            of the vermis and fastigial nucleus, but       ated with the tumor and its treatment.
from mutism follows the loss and regain       both the vermis and the cerebellar             Earlier studies of outcomes for patients
of simple orofacial movements as well as      hemispheres are involved in executive,         with brain tumors focused on survivor-
coordination of oral and pharyngeal           linguistic, and visual-spatial functions       ship and general intellectual function.
phases of swallowing [Dailey et al.,          [Schmahmann, 1991; Tavano et al.,              Improved survivorship and a growing
1995; Van Mourik et al., 1997]. How-          2007]. At this time, studies have not          collaboration between neurology, neuro-
ever, normal oral movements and swal-         clearly indicated how many patients            surgery, neuroradiology, and neuropsy-
lowing in some patients with CMS indi-        with mutism also have affective changes,       chology have led to more research on
cate that the problem is not purely           although the two are often described           treatment-related deficits in these chil-
motor [Dietze and Mickle, 1990]. It has       together in patients with CMS. Dem-            dren. Over the past 25 years, more than
also been suggested that CMS may rep-         onstrating that mutism rarely occurs           300 cases of mutism and associated be-
resent an extreme form of dysarthria          without affective changes would further        havioral and personality changes have
[Van Calenbergh et al., 1995]. Dysarth-       support a functional neuroanatomic             been reported after the removal of pos-
ria due to cerebellar infarction has been     overlap between these two features.            terior fossa tumors. Neuroradiographic
described for all vascular territories of            Most cases of postoperative CMS         studies have failed to demonstrate con-
the cerebellum, with the superior cere-       have occurred in the pediatric age             sistent preoperative or immediate post-
bellar artery being most often affected       group [Ildan et al., 2002]. Some studies       operative findings that predict who will
and the rostral paravermal area of the        have found younger age at diagnosis to         develop CMS. Although the syndrome
anterior lobe the most frequent lesion        be associated with poorer outcomes             is considered a postoperative complica-
site. FMRI data of healthy volunteers         [Packer et al., 1989; Mulhern et al.,          tion, a specific change in surgical tech-
show activation of this area during artic-    2001; Ronning et al., 2005], some              nique has not been proven. Given the
ulatory movements of the tongue and           studies have found young age to be             significant increase in survivorship of
orofacial muscles [Urban et al., 2003].       protective [Levisohn et al., 2000], and        patients with more aggressive resection
Recently, a more complex role in              others have found no association [Mab-         and the lack of evidence for a specific
speech production has been proposed.          bott et al., 2008]. These differences may      mechanism of injury, there are currently
Nonmotor pathways of speech subserve          be due to patient population differen-         no recommendations for changing sur-
the internal generation of words, word        ces. For instance, in Levisohn’s study,        gical technique. Future research needs
choice, speech rehearsal, and vocal into-     the majority of younger children non-          to address the seeming vulnerability to
nations used in speech production, and        medulloblastoma tumors and were                CMS in some patients. Perhaps func-
can be localized to the posterior lateral     therefore at lower risk for CMS. In a          tional MRI could shed light on whether
hemispheres of the cerebellum [Tohgi          review of pediatric studies supporting         differences in brain organization put cer-
et al., 1993]. The findings in CMS add        the role of the cerebellum in cognitive        tain individuals at risk for CMS. Such
to the growing body of evidence that          processes, Steinlin [2007] concludes that      knowledge could also help guide reha-
the cerebellum is involved in both coor-      earlier damage leads to more pro-              bilitation and counseling, as long-term
dination of orofacial movements and the       nounced problems. He hypothesizes              prognosis is currently poor. n
cognitive component of speech genera-         that the most important function of the
tion [Ozgur et al., 2006].                    cerebellum is learning, either during
        In cases of cerebellar dysfunction,   development or later on during a re-           REFERENCES
                                                                                             Akil H, Statham PF, G€tz M, et al. 2006. Adult
it is often difficult to differentiate the    habilitation process. The reason why                cerebellar mutism and cognitive-affective
contribution of the cerebellum from           mutism occurs in children and not typi-             syndrome caused by cystic hemangioblas-
other structures and connections in the       cally adults with CCAS may be related               toma. Acta Neurochir 148:597–598.

226                                                        Dev Disabil Res Rev         Cerebellar Mutism Syndrome           WELLS ET AL.
Al-Anazi A, Hassounah M, Sheikh B, et al. 2001.         Grill J, Viguier D, Kieffer V, et al. 2004. Critical          anatomy of single-word processing. Nature
      Cerebellar mutism caused by arteriovenous               risk factors for intellectual impairment in             331:585–589.
      malformation of the vermis. Br J Neurosurg              children with posterior fossa tumors: the role   Pinter JD, Eliez S, Schmitt JE, et al. 2001. Neuro-
      15:47–50.                                               of cerebellar damage. J Neurosurg 101(2                 anatomy of Down’s syndrome: a high-reso-
Andreasen NC, O’Leary DS, Cizadlo T, et al.                   Suppl):152–158.                                         lution MRI study. Am J Psychiatry 158:
      1995. PET studies of memory: novel versus         Huber JF, Bradley K, Spiegler BJ, et al. 2006.                1659–1665.
      practiced free recall of word lists. Neuro-             Long-term effects of transient cerebellar        Pollack IF. 1997. Posterior fossa syndrome. Int
      image 2:296–305.                                        mutism after cerebellar astrocytoma or                  Rev Neurobiol 41:411–432.
Aarsen F, VanDongen H, Paquier P, VanMourik                   medulloblastoma tumor resection in child-        Pollack IF, Polinko P, Albright AL, et al. 1995.
      M, Catsman-Berrevoets C. 1995. Mutism                   hood. Childs Nerv Syst 22:132–138.                      Mutism and pseudobulbar symptoms after
      and pseudobulbar symptoms after resection         Ildan F, Tuna M, Erman T, et al. 2002. The eval-              resection of posterior fossa tumors in children:
      of posterior fossa tumors in children: inci-            uation and comparison of cerebellar mutism              incidence and pathophysiology. Neurosurgery
      dence and pathophysiology. Neurosurgery                 in children and adults after posterior fossa            37:885–893.
      37:885–893.                                             surgery: report of two adult cases and review    Riva D, Giorgi C. 2000. The cerebellum contrib-
Aarsen F, VanDongen H, Paquier P, VanMourik                   of the literature. Acta Neurochir 144:                  utes to higher functions during develop-
      M, Catsman-Berrevoets C. 2004. Long-term                463–473.                                                ment: evidence from a series of children sur-
      sequelae in children after cerebellar astrocy-    Koh S, Turkel SB, Baram TZ. 1997. Cerebellar                  gically treated for posterior fossa tumours.
      toma surgery. Neurology 27:1311–1316.                   mutism in children: report of six cases and             Brain 123:1051–1061.
Bellebaum C, Daum I. 2007. Cerebellar involve-                potential mechanisms. Pediatr Neurol             Robertson PL, Muraszko KM, Holmes EJ, et al.
      ment in executive control. Cerebellum 6:                16:218–219.                                             2006. Incidence and severity of postoperative
      184–192.                                          Leiner HC, Leiner AL, Dow RS. 1986. Does the                  cerebellar mutism syndrome in children with
Chang YC, Huang CC, Huang SC. 1998. Volu-                     cerebellum contribute to mental skills?                 medulloblastoma: a prospective study by the
      metric neuroimaging in children with neu-               Behav Neurosci 100:443–454.                             Children’s Oncology Group. J Neurosurg
      rodevelopmental disorders—mapping the             Levisohn L, Cronin-Golomb A, Schmahmann JD.                   105:444–451.
      brain and behavior. Zhonghua Min Guo                    2000. Neuropsychological consequences of         Rnning C, Sundet K, Duc-Tnnessen B, Lundar
      Xiao Er Ke Yi Xue Hui Za Zhi 39:                        cerebellar tumour resection in children: cer-           T, Halseth E. 2005. Persistent cognitive dys-
      285–292.                                                ebellar cognitive affective syndrome in a               function secondary to cerebellar injury in
Cornish K, Kogan C, Turk J, Manly T, James N,                 paediatric population. Brain 123:1041–1050.             patients treated for posterior fossa tumors in
      Mills A, Dalton A. 2005. The emerging             Liu GT, Phillips PC, Molloy PT, et al. 1998. Vis-             childhood. Pediatr Neurosurg 41:15–21.
      fragile X permutation phenotype: evidence               ual impairment associated with mutism after      Sadeh M, Cohen I. 2001. Transient loss of speech
      from the domain of social cognition. Brain              posterior fossa surgery in children. Neuro-             after removal of posterior fossa tumors—one
      Cogn 57:53–60.                                          surgery 42:253–256.                                     aspect of a larger neuropsychological entity:
Dailey AT, McKhann GM II, Berger MS. 1995.              Mabbott DJ, Penkman L, Witol A. 2008. Core                    the cerebellar cognitive affective syndrome.
      The pathophysiology of oral pharyngeal                  neurocognitive functions in children treated            Pediatr Hematol Oncol 18:423–426.
      apraxia and mutism following posterior fossa            for posterior fossa tumors. Neuropsychology      Sagiuchi T, Ishii K, Aoki Y. 2001. Bilateral
      tumor resection in children. J Neurosurg                22:159–168.                                             crossed cerebello-cerebral diaschisis and
      83:467–475.                                       Middleton FA, Strick PL. 1994. Anatomical evi-                mutism after surgery for cerebellar medullo-
De Smet HJ, Baillieux H, Catsman-Berrevoets C,                dence for cerebellar and basal ganglia                  blastoma. Ann Nucl Med 15:157–160.
      et al. 2007. Postoperative motor speech pro-            involvement in higher cognitive function.        Schmahmann JD. 1991. An emerging concept.
      duction in children with the syndrome of                Science 266:458–461.                                    The cerebellar contribution to higher func-
      ‘cerebellar’ mutism and subsequent dysarth-       Mostofsky SH, Reiss AL, Lockhart P, et al. 1998.              tion. Arch Neurol 48:1178–1187.
      ria: a critical review of the literature. Eur J         Evaluation of cerebellar size in attention-      Schmahmann JD, Sherman JC. 1998. The cere-
      Paediatr Neurol 11:193–207.                             deficit hyperactivity disorder. J Child Neurol          bellar cognitive affective syndrome. Brain
Di Cataldo A, Dollo C, Astuto M, et al. 2001.                 13:434–439.                                             121(Part 4):561–579.
      Mutism after surgical removal of a cerebellar     Mulhern R, Palmer S, Reddick W, Glass J, Kun           Siffert J, Poussaint TY, Goumnerova LC, et al.
      tumor: two case reports. Pediatr Hematol                L, Taylor J, et al. 2001. Risks of young age            2000. Neurological dysfunction associated
      Oncol 18:117–121.                                       for selected neurocognitive deficits in                 with postoperative cerebellar mutism. J Neuro-
Dietze DD Jr, Mickle JP. 1990. Cerebellar mutism              medulloblastoma are associated with white               oncol 48:75–81.
      after posterior fossa surgery. Pediatr Neuro-           matter loss. J Clin Onc 19:472–479.              Steinbok P, Cochrane DD, Perrin R. 2003. Mutism
      surg 16:25–31.                                    Nandagopal R, Krishnamoorthy SG. 2004. Tran-                  after posterior fossa tumour resection in chil-
Eckert MA, Leonard CM, Richards TL, et al.                    sient mutism due to posterior circulation in-           dren: incomplete recovery on long-term fol-
      2003. Anatomical correlates of dyslexia:                farction. Neurol India 52:510–511.                      low-up. Pediatr Neurosurg 39:179–183.
      frontal and cerebellar findings. Brain 126        Nicolson R, Fawcett AJ, Dean P. 2001. Dyslexia,        Steinlin M. 2007. The cerebellum in cognitive
      (Part 2):482–494.                                       development and the cerebellum. Trends                  processes: supporting studies in children.
Ersahin Y, Mutluer S, Cali S, et al. 1996. Cere-             Neurosci 24:515–516.                                    Cerebellum 6:237–241.
      bellar mutism: report of seven cases and          Ozgur BM, Berberian J, Aryan HE. 2006. The             Stoodley CJ, Fawcett AJ, Nicolson RI, et al. 2006.
      review of the literature. Neurosurgery 38:              pathophysiologic mechanism of cerebellar                Balancing and pointing tasks in dyslexic and
      60–65.                                                  mutism. Surg Neurol 66:18–25.                           control adults. Dyslexia 12:276–288.
Ersahin Y, Yararbas U, Duman Y, et al. 2002. Sin-       Ozimek A, Richter S, Hein-Kropp C, et al. 2004.        Tavano A, Grasso R, Gagliardi C. 2007. Disorders
      gle photon emission tomography following                Cerebellar mutism—report of four cases.                 of cognitive and affective development in
      posterior fossa surgery in patients with and            J Neurol 251:963–972.                                   cerebellar malformations. Brain 130(Part 10):
      without mutism. Childs Nerv Syst                  Packer RJ, MacDonald T, Vezina G. 2008. Cen-                  2646–2660.
      18:318–325.                                             tral nervous system tumors. Pediatr Clin         Timmann D, Daum I. 2007. Cerebellar contribu-
Fujisawa H, Yonaha H, Okumoto K, et al. 2005.                 North Am 55:121–145.                                    tions to cognitive functions: a progress
      Mutism after evacuation of acute subdural         Packer RJ, Sutton LN, Atkins TE, et al. 1989. A               report after two decades of research. Cere-
      hematoma of the posterior fossa. Childs                 prospective study of cognitive function in              bellum 6:159–162.
      Nerv Syst 21:234–236.                                   children receiving whole-brain radiotherapy      Tohgi H, Takahashi S, Chiba K, et al. 1993. Cer-
Gelabert-Gonzlez M, Fernndez-Villa J. 2001.
                a              a                              and chemotherapy: 2-year results. J Neuro-              ebellar infarction. Clinical and neuroimaging
      Mutism after posterior fossa surgery. Review            surg 70:707–713.                                        analysis in 293 patients. The Tohoku Cere-
      of the literature. Clin Neurol Neurosurg          Palmer SL, Reddick WE, Gajjar A. 2007. Under-                 bellar Infarction Study Group. Stroke
      103:111–114.                                            standing the cognitive impact on children               24:1697–1701.
Giedd JN, Blumenthal J, Molloy E, et al. 2001.                who are treated for medulloblastoma. J           Turgut M. 1998. Transient ‘‘cerebellar’’ mutism.
      Brain imaging of attention deficit/hyperac-             Pediatr Psychol 32:1040–1049.                           Childs Nerv Syst 14:161–166.
      tivity disorder. Ann N Y Acad Sci                 Papavasiliou AS, Kotsalis C, Trakadas S. 2004.         Turgut M. 2007. Re: the pathophysiologic mech-
      931:33–49.                                              Transient cerebellar mutism in the course of            anism of cerebellar mutism. Surg Neurol
Glauser TA, Packer RJ. 1991. Cognitive deficits               acute cerebellitis. Pediatr Neurol 30:71–74.            68:117.
      in long-term survivors of childhood brain         Petersen SE, Fox PT, Posner MI. 1988. Positron         Turgut M. 2008. Cerebellar mutism. J Neurosurg
      tumors. Childs Nerv Syst 7:2–12.                        emission tomographic studies of the cortical            Pediatrics 1:262.

Dev Disabil Res Rev              Cerebellar Mutism Syndrome               WELLS ET AL.                                                                        227
Urban PP, Marx J, Hunsche S, et al. 2003. Cere-          terior fossa surgery in children. Neurosur-        cases. Dev Med Child Neurol 39:686–
     bellar speech representation: lesion topogra-       gery 37:894–898.                                   690.
     phy in dysarthria as derived from cerebellar    Van Dongen HR, Catsman-Berrevoets CE, van         Wells EM, Khademian ZP, Walsh KS, et al. Neu-
     ischemia and functional magnetic resonance          Mourik M. 1994. The syndrome of ‘cere-             roradiographic features of patients with the
     imaging. Arch Neurol 60:965–972.                    bellar’ mutism and subsequent dysarthria.          Cerebellar Mutism Syndrome. Ann Neurol
Vadeinse D, Hornyak JE. 1997. Linguistic and             Neurology 44:2040–2046.                            (in press).
     cognitive deficits associated with cerebellar   Van Mourik M, Catsman-Berrevoets CE, van          Zang YF, He Y, Zhu CZ, et al. 2007. Altered
     mutism. Pediatr Rehabil 1:41–44.                    Dongen HR, et al. 1997. Complex                    baseline brain activity in children with
Van Calenbergh F, Van de Laar A, Plets C, et al.         orofacial movements and the disappearance          ADHD revealed by resting-state functional
     1995. Transient cerebellar mutism after pos-        of cerebellar mutism: report of five               MRI. Brain Dev 29:83–91.

228                                                                 Dev Disabil Res Rev          Cerebellar Mutism Syndrome             WELLS ET AL.