Daniel

 CHAPTER 5.3 CHILDHOOD-ONSET

SCHIZOPHRENIA AND OTHER EARLY-ONSET

PSYCHOTIC DISORDERS

ANNA E. ORDÓÑEZ AND NITIN GOGTAY

BACKGROUND

Psychotic disorders are rare in children although transient psychotic experiences are

more common in otherwise healthy children than generally recognized (1–3). As is

often the case with other very early-onset illnesses, psychotic disorders in children

are usually more severe than their adult counterparts (4), and the disruption of

cognitive and social development as well as the burden to the family can be

devastating. Systematic research in this area was initially limited by diagnostic

uncertainty and the general lack of knowledge about the psychotic processes in

children (5).

History of Childhood-Onset Psychoses

Although the existence of childhood schizophrenia was recognized early in the

twentieth century (6), the term psychosis was used so broadly in children that a

spectrum of behavioral disorders and autism were grouped together under the

category of childhood schizophrenia (7). The landmark studies of Kolvin (8) first

established the clinical distinction between autism and other psychotic disorders of

childhood. However, even today high rates of initial misdiagnosis remain due to

symptom overlap, particularly for mood disorders (9–11), and the presence of

transient hallucinations and delusions in nonpsychotic pediatric patients (3,12,13).

Overall, hallucinations and delusions appear more prevalent in children and early

adolescents than in older adolescents (14,15), with a recent meta-analysis of

population studies reporting median prevalence of psychotic symptoms among

children ages 9 to 12 years of 17%, and among adolescents 13 to 18 years of 7.5%.

Nonetheless, while the occurrence of psychotic-like experiences in many children

and early adolescents will have short-term discontinuation, older age (14), greater

severity, and frequency (12) appear associated to persistence of symptoms.

Furthermore, psychotic symptoms in adolescents can be important indicators of risk

for a range of psychotic (13) and nonpsychotic mental disorders (14–16) as well as

conferring greater risk of poorer outcomes (17). In summary, while prevalent and

usually transient in nonclinical populations particularly at earlier ages, psychotic

symptoms in childhood and adolescents can be predictive of psychopathology later in

life (16,18), suggesting that these symptoms probably exist as a continuous phenotype

rather than an all-or-none phenomenon (19).

Childhood-Onset Schizophrenia—Diagnosis, Clinical

Presentation, and Differential Diagnoses

Given the relatively high rate of hallucinations in children, with prevalence estimates

in large population studies of 8% (20), and reports as high as 21.3% in some

community samples (21), the distinction of pathologic psychotic symptoms in

childhood is very important. For instance, 28% to 65% of children between the ages

of 5 and 12 years of age report experiencing imaginary friends (22) which could be

inappropriately misinterpreted as pathologic. Similarly, hallucinations associated

with the periods of transition from wakefulness to sleep (hypnagogic) or from sleep

to wakefulness (hypnopompic), are frequent in childhood and decline with age (23).

It is noteworthy that some cases of sleep-associated hallucinations can be indicative

of sleep disorders such as narcolepsy (24).

Although rare, with an incidence in the National Institute of Mental Health

(NIMH) cohort of less than 0.04% (25), Kolvin’s work established that children can

be diagnosed with unmodified criteria for schizophrenia (8,25). Childhood-onset

schizophrenia (COS) shows a pattern similar to that of poor-outcome adult cases, and

the psychosis of COS can usually be distinguished by its severe and pervasive nature

and its nonepisodic, unremitting course (25). Additionally, these children show

poorer premorbid functioning in social, motor, and language domains, learning

disabilities, and disruptive behavior disorders (26–28), and although not reported in

studies of the premorbid history of adult-onset schizophrenia (AOS) (29,30),

transient autistic symptoms such as hand flapping and echolalia in toddler years are

common (26), probably reflecting more compromised early brain development.

Because COS is rare, it must be distinguished from several childhood conditions

that can manifest with psychotic symptoms and/or deterioration in function:

1. Affective disorders: Hallucinations are relatively common in pediatric bipolar

disorder and major depression (9,17,31,32). However, the psychotic symptoms

in these conditions tend to be mood congruent (33) and follow-up studies on this

population generally suggest a less disabling long-term course in a large number

of cases (34,35).

2. Psychosis due to medical conditions such as migraine (36), inborn errors of

metabolism (37), and substance abuse disorders (38) should be carefully ruled

out.

3. Autism spectrum disorders and childhood disintegrative disorder can often be

mistaken for psychosis, as they show severe impairment in reciprocal

communication, social interactions, and odd stereotyped behaviors (39,40).

4. Conduct disorder and various other behavioral disturbances can be associated

with hallucinations (41,42).

5. Atypical psychosis is an important differential diagnosis and is described in

detail below.

Atypical Psychosis

A sizeable, heterogeneous group of children referred to the NIMH COS study over

the past 20 years had transient psychotic symptoms and multiple developmental

abnormalities, but were not adequately characterized by existing DSM-IV categories

and were given a provisional diagnosis of “multidimensionally impaired (MDI)”

(43–45). In the DSM nosology, these patients might be considered as having either

psychosis NOS or mood disorder NOS.

The MDI group, although showing similarities with COS, has distinct features

which were used as the operational diagnostic criteria by the NIMH group (43,44)

including:

1. Brief, transient episodes of psychosis and perceptual disturbance, typically in

response to stress (as opposed to the pervasive hallucinations/delusions in

COS)

2. Nearly daily periods of emotional lability disproportionate to precipitants

3. Impaired interpersonal skills despite the desire to initiate peer friendships

(distinction from COS)

4. Cognitive deficits as indicated by multiple deficits in information processing

5. No clear thought disorder (although this can be difficult to define clinically,

particularly in the presence of a communication disorder)

6. Comorbid ADHD

The MDI children resemble syndromes such as the borderline syndrome of

children, or the multiple complex developmental disorder (MCDD) (45–48).

However, these syndromes have more predominant symptoms of autism spectrum

disorders; greater evidence of a formal thought disorder; and onset before age 5 (as

opposed to about age 8 for the MDI group) (44,45,49,50).

Initially, the neuropsychologic test profiles, smooth pursuit eye movement (SPEM)

abnormalities, and familial risk factors suggested that some of the MDI children fell

within the schizophrenia spectrum (44), but the MDI cohort appears to have a

distinct long-term clinical course, and none have progressed to schizophrenia (51).

The NIMH COS Experience

The NIMH COS study has been ongoing since 1990. Inclusion criteria for the study

are as follows: onset of psychosis before 13, premorbid IQ of 70 or above, and

absence of significant neurologic disorder. In the last 24 years, over 2,000 charts

have been reviewed, of which 80% are declined from further consideration as they

fail to meet criteria for COS. About 400 children have been screened in person, of

whom about 60% receive other psychiatric diagnoses, such as affective disorders,

anxiety, or behavioral disorders. About 250 children who appeared likely to meet

criteria for COS have been admitted to the research unit and undergone a complete

medication washout followed by 1 to 3 weeks’ drug-free inpatient observation. An

additional 40% did not meet criteria for COS, most frequently because a diagnosis of

affective disorder was made (5,43,52). A 4- to 6-year follow-up study of 33 of the

ruled-out cases indicated good stability of the alternative diagnoses and confirmed

the absence of schizophrenia (53).

PHENOMENOLOGY AND NEUROBIOLOGY OF COS

Premorbid Development

A striking phenomenologic feature of COS relative to AOS appears to be the higher

rates of early language, social, and motor developmental abnormalities, possibly

reflecting greater impairment in early brain development. In the NIMH COS sample,

premorbid development (defined as development prior to 1 year before psychosis

onset and assessed using the Cannon-Spoor Premorbid Adjustment Scale (PAS) ( 54)

and the Hollis premorbid development scale (27) and social, speech and language

impairments were clearly impaired (55,56), as has been previously observed by

other independent research centers (27,28,57,58).

Risk Factors

Since COS represents a more severe phenotype of schizophrenia than AOS, our

initial hypothesis was that most risk factors identified in adult studies would be more

striking in our very early-onset cases.

Parental Age and Obstetric Complications

AOS studies suggest associations of the illness with advanced paternal age (59–63),

raising the possibility of increased de novo mutations in the paternal germ cells, and

also increased rates of maternal/fetal obstetric complications (64–66). The NIMH

COS cohort showed no correlation with maternal or paternal age (67). In our cohort,

we compared the obstetric records of 60 children with COS and 48 healthy siblings

using the Columbia Obstetrics Complication Scale (68), a comprehensive

measurement scale consisting of 37 variables. Contrary to our hypothesis, the

incidence of obstetric complications in COS patients did not differ from that for the

healthy sibling control group (69).

Eye Tracking

SPEM abnormalities have been reported in 25% to 40% of first-degree relatives of

schizophrenia patients (70). Other studies have suggested more striking abnormalities

in COS than in AOS, with a bilineal pattern of inheritance (71). Our group compared

70 COS parents, 64 AOS parents, and 20 COS siblings to separate matched control

groups and found that the effect sizes for SPEM abnormalities were higher for COS

than for AOS relatives, indicating that genetic factors underlying eye-tracking

dysfunction may be more salient for COS (72).

Familial Schizophrenia Spectrum Disorders

Schizophrenia spectrum disorders consist of schizophrenia and schizoaffective

disorders on axis I, and schizotypal, paranoid, and schizoid personality disorders on

axis II (73). A prior study by Asarnow et al. ( 74) showed higher rates of

schizophrenia spectrum diagnoses for COS relatives than for relatives of patients

with attention-deficit hyperactivity disorder (ADHD) or community controls.

Similarly, our analyses of 97 parents of COS patients, 97 parents of AOS patients

and matched community controls, found a higher rate of familial schizophrenia

spectrum disorders in COS than AOS, with the lowest rate in community controls.

This observation further supported the continuity between COS and AOS, and the

salience of familial/genetic risk in COS (75).

Familial Neurocognitive Functioning

It is well documented that subtle cognitive deficits, including abnormalities of

attention (76,77), executive functioning (78,79), spatial working memory (80), and

verbal memory (81,82) are seen in healthy relatives of patients with AOS (83–85).

Deficits in auditory attention, verbal memory, and executive functioning are generally

considered consistent (86,87), although it is unclear whether these represent an

underlying global cognitive deficit or each deficit represents a discrete

endophenotype transmitted in families of patients with schizophrenia (78,86). When

we compared neuropsychologic deficits in 67 parents and 24 full siblings of COS

patients to matched community controls in the Trail Making Tests A and B and the

Wechsler Intelligence Scale-Revised Digit Span and Vocabulary, COS siblings had

significantly poorer performance than community controls, although the rates of

neuropsychologic abnormalities for COS were not significantly higher than for AOS

(88).

Pervasive Developmental Disorder and COS

The diagnosis of autism or pervasive developmental disorder (PDD) has been raised

early in the development in our cases (n = 28 of 97, 28%) (26) and several studies

have claimed that autism per se might be a risk factor for later psychosis (89–92).

Premorbid PDD may be a nonspecific manifestation of impaired neurodevelopment,

and also may provide an independent additive risk factor for COS. We ascertained

the premorbid diagnosis of past or current autism or PDD in 97 COS probands in

whom the diagnosis of PDD was made according to DSM-IV criteria (American

Psychiatric Association, 1994) based on early chart reviews and clinical interviews

of patients and parents. The diagnoses of PDD was made in cases in which there was

a clear clinical history of symptoms of autistic/PDD spectrum prior to the onset of

psychosis that were still observable at the time of the NIH evaluation.

Twenty-eight (29%) of our COS patients had a lifetime diagnosis of PDD: 3 met

criteria for autism, 1 for Asperger disorder, and 24 for PDD NOS ( 26). Premorbid

social impairment was the most common feature for COS-PDD subjects; the PDD

group did not differ significantly from the rest of the COS sample with respect to

SES, age of onset, IQ, or severity as measured by the Global Assessment of Severity

(GAS). However, similar to demographic characteristics seen in the literature for

children with PDD, the COS patients with PDD were predominantly male (p = 0.04)

and non-African American (p = 0.02) (26). In an earlier study, our group also failed

to find any differences in baseline or 2- to 6-year clinical outcome measures,

parental SES, response to medications, and rate of familial schizotypy (56).

Furthermore, there was no difference between PDD and non-PDD groups with

respect to initial brain magnetic resonance imaging (MRI) measures after controlling

for gender, although the rate of gray matter (GM) loss appeared to be greater for

PDD (n = 12) than for the non-PDD (n = 27) subgroup (−19.5 ± 11.3 mL/yr vs. −9.6

± 15.3 mL/yr; p = 0.05) (56). These results may indicate that PDD in COS may be a

nonspecific marker of more severe early abnormal neurodevelopment. However,

siblings of the PDD-COS patients had significantly higher scores on the autism

screening questionnaires, and 2 of 12 (17%) siblings of PDD patients were

diagnosed with autism, a total rate similar to that seen for sibling of autistic patients

(4.9%) (93), which may still imply a familial–genetic connection between COS and

autism.

Neurocognitive Functioning in COS Patients

Neuropsychologic function in COS outpatients has been studied by Asarnow et al.

(94–96). While rote language skills and simple perceptual processing are not

impaired, these children perform poorly on tasks involving fine motor coordination,

attention, short-term and working memory (97). Evoked potential studies show

diminished amplitude of brain electrical activity during these tasks, suggesting that

allocation of necessary attentional resources is deficient, which is also shared by

adults with schizophrenia (96). It is generally established for adult schizophrenia that

cognitive function deteriorates at onset of psychosis but remains stable afterward

(98–100). Our earlier study had shown that children with COS (n = 27) as well as

those with MDI (n = 24) share similar deficits in attention, learning, and abstraction,

resembling the pattern in adult patients with schizophrenia (101). A subsequent

analyses of 71 COS patients where preadmission IQ data were also available from

medical and school records for a subgroup (n = 27), pre- and postpsychosis decline

in IQ was noted as for adults, but postpsychotic cognitive function for up to 8+ years

did not show continued decline (Fig. 5.3.1). Thus, in spite of greater severity and

generally poor clinical outcome, there was no evidence for a longer-term

degenerative cognitive process in COS, at least through early adulthood (102).

Comorbid Disorders

Comorbid psychiatric disorders, particularly DSM defined mood, anxiety, and

substance abuse disorders, often coexist with schizophrenia and can significantly

alter the presentation, clinical course, or prognosis of the illness (103–106). As the

symptom manifestations of these disorders can also be part of (or masked by) the

symptoms of the primary illness, the diagnoses of independent axis I conditions are

often ignored (107–109). Furthermore, medical comorbidities in schizophrenia are

also frequent (110), including cardiovascular disease (due to tobacco, diabetes,

obesity, and increased lipids), HIV, and infectious hepatitis ( 111). We analyzed the

rate of coexistent axis I diagnoses for 76 COS cases at the time of first NIMH

admission, and correlated the comorbid diagnoses with age of onset, ratings of

illness severity, familiarity, and premorbid development.

As seen with AOS, the most frequent comorbid diagnosis at NIMH screening was

depression (54%), followed by obsessive-compulsive disorder ([OCD] 21%),

generalized anxiety disorder ([GAD] 15%), and ADHD (15%). The rate of “any”

anxiety disorder (GAD, OCD, separation anxiety, PTSD, and panic disorder

combined) at screening was 42%. In general, comorbid diagnoses were independent

of other illness indices, but comorbid depression correlated with poorer GAS scores

(p = 0.01), and presence of an anxiety disorder only predicted anxiety at 4-year

follow-up (p = 0.05). No other psychiatric diagnoses showed correlations with any

clinical measures, and there were no significant associations between comorbid

diagnoses and IQ, familiarity, medication status, premorbid functioning, or age of

onset at psychosis. Interestingly, no “current” comorbid depression was seen at the 4-

year follow-up for a subgroup of 28 subjects for whom there was complete

diagnostic information available, possibly due to our high use of antidepressant

treatment (45%). In contrast, anxiety disorders remained highly comorbid despite

adjuvant anxiety medication use, suggesting either the refractory nature of these

conditions, or a close association with core schizophrenia pathology.

FIGURE 5.3.1. Full scale IQ measures for 70 COS children plotted

before (n = 21) and after (n = 70) the onset of psychosis, and also

before (n = 56; obtained from prior charts) and after (n = 70) the

children were admitted to the NIMH study. Although the children

show significant decline in full scale IQ after the onset of psychosis,

there is no significant long-term decline over next 14 years.

BRAIN IMAGING: STRUCTURAL, FUNCTIONAL, AND

POSITRON EMISSION TOMOGRAPHY

The majority of the imaging studies of COS come from the NIMH sample, with more

recent contributions from other groups (112–115). Using both cross-sectional and

longitudinal data, and rapidly developing state-of-the-art brain mapping methods,

neuroimaging in COS has addressed some broad questions about brain development

in COS and schizophrenia in general.

Structural Neuroimaging

A fundamental question in brain imaging studies of any illness is whether there are

overall differences in brain size. Studies in AOS have documented decreased

intracranial volume (116,117), longitudinal reduction in total cerebral volume (118),

and ventricular enlargement (119). It is also predicted that brain growth in AOS is

stunted even before the onset of illness (116). Similarly, in COS patients, overall

brain volume at initial scan is smaller and is followed by a progressive decline in

volume during adolescence (120). Patients with COS have also been found to have

larger ventricular volume (121) as well as greater progressive increase in

ventricular size compared to healthy controls (120,122). Understanding how these

brain changes are meaningfully related to clinical features remains an important

unanswered question, but collective research points toward the significance of key

findings, which are highlighted in the following sections.

Cortical Gray Matter Thickness

Progressive cortical GM loss in COS was first described by Thompson et al. in 2001

(123). This study demonstrated a dynamic wave of GM loss, which started in the

parietal and motor cortices and advanced into the superior frontal, dorsolateral

prefrontal, and temporal cortices (including the superior temporal gyri). While

temporal and dorsolateral prefrontal cortex deficits were among the most severe,

they began in late adolescence and were observed only after the onset of psychotic

symptoms. The progressive deterioration in GM also correlated with overall

deterioration in global functioning. Seeking to determine whether psychosis itself

was associated with GM loss, the study included a comparison group of participants

with atypical (nonschizophrenic) psychosis, matched for IQ and medications, in

addition to healthy volunteers. This atypical psychosis group showed subtle but

significantly greater GM loss than healthy volunteers, pointing to a successively

increasing rate of GM loss, with COS patients experiencing the greatest loss. In

2004, prompted by the Thompson study, using a similar analysis we evaluated

cortical maturation in typically developing children for prospective cortical changes

between the ages of 4 to 21 years (124). This study highlighted the timelines of

cortical maturation, and importantly, it suggested that COS neurodevelopment

appeared to be an exaggeration of normal GM loss and maturation patterns, possibly

indicating a loss of inhibitory regulation of the development process (Fig. 5.3.2).

These initial observations generated an important question: Does the GM loss

persist into early adulthood and if so, does it continue at the same rate? A

longitudinal study addressed this question by following COS patients and controls

into adulthood (125). This study found a 7.5% difference in mean cortical thickness

(MCT) (p = 0.001) between COS patients (n = 70, ages 7 to 26) and age-matched

healthy controls (n = 72), as well as progressive cortical thinning in the parietal,

frontal, and temporal regions, although the parietal thinning normalized by early

adulthood (125). These results are continuous with findings from AOS patients,

which demonstrate cortical thinning in the frontal and temporal cortices only

(126,127). Together, these findings established that the profound GM thinning in

adolescence appears to slow down as the children mature. Whether this lessening

rate is part of the natural course of the disease, a resilience process, or due to

medical treatment remains difficult to conclude.

It has also been established that COS patients do not differ from healthy controls

with regard to sex differences in cortical thickness (128), or cross-sectional or

longitudinal developmental changes in asymmetry (129). Cortical thickness deficits

in COS are also largely uninfluenced by clozapine versus olanzapine intake, aside

from a small area of the right prefrontal cortex (130). These findings are consistent

with AOS, in which age, dose, or type of antipsychotic medication is not significantly

linked to changes in cortical thickness (127).

Subcortical Structures

Hippocampus Critical in learning and memory, the hippocampus has been a structure

of significant interest in schizophrenia, where cognitive deficits remain a primary

feature of the disease. Deficits in hippocampal volume are well documented in AOS

bilaterally (131) and is suggested in COS (120,132).

Prospective studies (133,134) have demonstrated fixed longitudinal volumetric

deficits in COS patients compared to controls. Of these, the study with the largest

sample (134) (89 COS patients, 78 siblings, and 79 controls) corroborated prior

findings (120,133) that suggest that the hippocampal deficits in COS are significant

but do not vary over time. These findings are consistent with the animal model of

schizophrenia suggested by Lipska et al. in 1993 (135), in which an early fixed

deficit in the hippocampus, which is relatively quiescent during the first years of

development, manifests later during periods of increased stress or with exposure to

particular substances.

Anatomically, subregional shape abnormalities of hippocampus have also been

described in AOS (136,137) and COS (138). Bilateral inward deformations of the

anterior hippocampus have been found in COS (138), and greater differences have

been associated with increased symptom severity. The affected hippocampal regions

are commonly associated with hippocampal CA1 pyramidal neurons, whose

migration is disrupted by genetic differences associated with schizophrenia (e.g.,

disruption of the Disc1 gene) (139). The CA1 neurons serve as a connection between

the hippocampus and the prefrontal cortex, which is also widely implicated in

schizophrenia (140). While the exact involvement of hippocampal abnormalities in

COS is unclear, it is possible that abnormal neurodevelopment in these

interconnected regions holds further information regarding symptom development Cerebellum The cerebellum has a highly heritable development (141), may

contribute to higher cognitive functions (142), and it is a potential key site of

dysregulated circuitry in schizophrenia (143). In line with this, cerebellar deficits

have been linked to COS (144). An early cross-sectional study of COS patients

between the ages of 9 and 18 demonstrated decreased cerebellar volume relative to

controls in the vermis (11.7% smaller), midsagittal inferior posterior lobe area

(10.9% smaller), and midsagittal inferior posterior lobe (8.9% smaller) (145).

Studies of AOS have reported a progressive decline in cerebellar volume (146), or a

smaller cerebellar volume during the first episode in AOS (147).

In a prospective study of COS patients (n = 94) ages 6 to 29 and their

nonpsychotic siblings (n = 80), COS subjects had smaller bilateral anterior lobes and

anterior and total vermis volumes compared to controls (148). Siblings did not differ

from healthy controls initially, but demonstrated decreased cerebellar volume over

time in the total and right cerebellum, left inferior posterior, left superior posterior,

and superior vermis (148). The presence of cerebellar deficits in healthy siblings of

COS patients, as well as the presence of abnormalities in AOS patients during the

first episode, suggests that the cerebellar trajectory described is likely related to a

genetic risk for schizophrenia.

White Matter The dysconnectivity hypothesis of schizophrenia first proposed that

aspects of the illness could be due to abnormal (increased or decreased) connectivity

between brain regions as opposed to localized abnormalities within regions (149).

Studies of white matter provide initial insights into the dysconnectivity hypothesis.

White matter abnormalities have been demonstrated in both AOS and COS patients.

Using tensor-based morphometry, we demonstrated that teenage COS patients have a

slower rate of white matter growth per year, particularly in the right hemisphere, and

that growth deficits are associated with lower functioning in terms of the GAS (150).

AOS patients show similar white matter deficits longitudinally (117).

Diffusion tensor imaging (DTI), which examines directional diffusion of water in

the brain to infer the structural integrity of neural fibers, has led to less conclusive

results. In the largest study of DTI in COS to date, examining eleven regions of

interest, we found decreased white matter integrity in the bilateral cuneus, a portion

of the occipital lobe (151). In another DTI study, COS patients exhibited a decrease

in white matter integrity in relation to their level of linguistic impairment (152).

Overall though, DTI research in COS is limited, and inconsistent with the frontal

abnormalities demonstrated by AOS DTI research (153); more observations with

larger samples and better image resolution are likely required to provide clearer

insight into potential white matter alterations in COS.

Corpus Callosum The majority of patients with schizophrenia manifest neurologic

soft signs that include errors in sensory integration, motor coordination, and

inhibition (154). These processing deficits, which may arise partly from decreased

interhemispheric neural communication, have been associated with irregularities in

the corpus callosum (155) and meta-analyses have generally established a reduced

midsagittal cross-sectional area in the corpus callosum (156,157). First episode

patients showed decreased corpus callosum areas, while chronic patients were more

likely to demonstrate an increase in the area (156). The only longitudinal study of the

corpus callosum in AOS schizophrenia patients demonstrated progressive decline in

callosal size, with poor-outcome patients showing more pronounced decline (158).

The largest to-date longitudinal study in COS (159) (n = 98), their siblings (n =

71), and healthy controls (n = 100) found no differences in total corpus callosum or

in the area of any subregion cross-sectionally or in developmental trajectories of any

measurement of corpus callosum area or volume (159). Similarly, a study examining

healthy siblings of COS patients found no significant differences in corpus callosum

area (159). These data suggest that the behavioral deficits in sensorimotor integration

in COS may originate not in callosal connections but in the interaction of multiple

networks, possibly reflecting a dysfunction in predictive processing (160–162).

Functional Magnetic Resonance Imaging

Functional imaging studies are difficult to conduct in the COS due to illness severity,

which make it difficult for patients to perform tasks in the scanner and often introduce

behavioral confounds. In a unique fMRI study of language processing in COS,

Borofsky et al. (163) found that COS patients have overall reduced activity

compared to healthy controls during both semantic and syntactic language processing

tasks. The differences in activation were not related to performance on the task, as

there were no group differences in success rate.

Resting state fMRI (R-fMRI) studies have been relatively more feasible in COS.

R-fMRI has been used to examine dysconnectivity in AOS using various methods,

and support the dysconnectivity model of schizophrenia (164). To date, the four

published studies of R-fMRI in COS are from the NIMH sample (165–167) and

demonstrate decreased local connectivity strength in COS that is partially balanced

by increased global network efficiency relative to healthy controls (166,167).

Furthermore, we found widespread decreased functional correlations in COS

patients compared to healthy volunteers, in brain regions that organized into a socialcognitive

and a sensorimotor processing network. These findings emphasized the role

of dysfunctional integration of these two major systems, and importantly, found that

resting state alterations were linked to behavioral symptoms. Specifically, decreases

in functional connectivity across these two networks in COS were related to the

severity of positive symptoms, while connectivity decreases within the socialcognitive

network related to the severity of negative symptoms (165). These results

are consistent with the proposal that dysfunctional network interactions play a crucial

role in the disease, particularly with respect to sensorimotor integration and

predictive processing.

Healthy Siblings

Studying nonsymptomatic or healthy siblings of patients with heritable illnesses

improves understanding of the contribution of genetic background to an illness state

versus an illness predisposition or trait. Nonpsychotic biologic full siblings of COS

patients, share about 50% of their genetic material and can add valuable context to

neuroimaging findings (168). In the most simplified form, consistent phenotypic

differences between patients and their healthy biologic full siblings suggest that the

patient phenotype is related to the expression of symptoms (disease state), assuming

that the healthy sibling group and unrelated healthy control group do not differ. On the

other hand, similarities between healthy and ill siblings, despite the difference in

their health status, can relate phenotypes to genetic vulnerability that may not

contribute directly to symptoms (disease trait). Due to the early-onset nature of COS,

siblings of COS patients enter studies at early ages and thus provide a unique

opportunity to address questions of state versus trait within a neurodevelopmental

window.

State Markers

State markers are characteristics unique to the expression of symptoms that do not

occur in unrelated healthy controls or healthy siblings of patients despite their

predisposition for the illness. Because these differences occur despite similarities in

genetic background between patients and their unaffected siblings, they are assumed

to be associated with the development of the disease, for example, psychosis in the

case of schizophrenia. The brain area most consistently identified as a state marker

for schizophrenia through COS studies is the hippocampus, which demonstrates

volumetric deficits, that are fixed over time, in patients but not their nonpsychotic

siblings (134).

Trait Markers

Trait markers are defined as phenotypes that may be related to predisposition for the

illness, but may not be directly related to the disease. Nonpsychotic siblings of COS

patients share a number of characteristic abnormalities with their affected relatives,

including differences in cerebral volume, GM thickness, and white matter growth

(169–171). Two nonoverlapping studies of unaffected siblings (n = 52; ages 8 to 28

and n = 43, ages 5 to 26 years) versus healthy controls (n = 52 and n = 86)

demonstrated no significant difference in MCT between the two groups, however the

sibling group had a pattern of early restricted GM loss (171,172). Both studies

examined cortical thickness in siblings by region, revealing GM deficits in

prefrontal, temporal, and parietal areas during early life. Each of these deficits,

though initially similar to those seen in young COS patients, were undetectable by

adulthood (171,172) (Fig. 5.3.3). This deficit “normalization,” which is likely to be

an age-specific trait marker, may also help explain inconsistent results from studies

of cortical thickness in healthy siblings of AOS patients (173) where the average age

of sibling sample is much older and past the developmental window. Furthermore,

siblings of COS patients have also been shown to exhibit deficits in white matter

growth during adolescence at early stages in life, however the growth rate normalizes

by adulthood in healthy siblings, suggesting that white matter growth may also

represent a trait marker (170).

FIGURE 5.3.3. Cortical GM development in healthy COS siblings (n

= 56; 100 scans) compared with age- and sex-matched normal

volunteers (n = 56; 100 scans) between 6 and 30 years. Analyses

were done using automated cortical thickness measure across

40,000 cortical points over the entire cortex using mixed-effect

regression model analyses which allowed using cross-sectional as

well as longitudinal scans. Color bars, which can be observed in

color in the original paper or in the online figure, represent tstatistics

where t > 2 (adjusted for multiple comparisons, using

false discovery rate at t = 2) indicates significant loss of GM in

healthy siblings compared to controls at that cortical point. GM

differences at various ages were obtained by age recentering the

data at that age. Healthy siblings showed GM loss in prefrontal and

temporal cortices in early ages, but the GM differences were

normalized by age 24 years. Images are shown both with and

without adjusting for mean cortical thickness (MCT). (From Gogtay

N, Greenstein D, Lenane M, et al.: Cortical brain development in

nonpsychotic siblings of patients with childhood-onset

schizophrenia. Arch Gen Psychiatry 64(7):772–780, 2007.)

To date, a single task-based fMRI study has been reported in a COS sibling

population which demonstrated that healthy siblings of COS patients showed aberrant

frontal and striatal activation relative to healthy controls during a cognitive skill

learning task (174). If feasible, it would be interesting to conduct a similar study in

COS patients, in order to understand to what extent these findings may represent a

functional endophenotype.

To our knowledge, no R-fMRI studies have been published describing functional

connectivity in healthy siblings of COS patients although preliminary findings from

our unpublished data suggest they are intermediate, similar to that seen for structural

abnormalities, suggesting a potential functional trait marker.

Genetic Studies

Schizophrenia is a complex disorder with a heterogeneous phenotype in which most

likely numerous genetic and environmental exposures are involved (175). While

inheritance patterns vary, a strong genetic inheritance is supported by twin, family,

and adoption studies (175,176). Observations across pediatrics and medicine suggest

that early-onset cases may have more salient genetic causes (4,177,178), and as with

breast cancer (179), Alzheimer disease (180), and type II diabetes (181), the familial

risk for schizophrenia spectrum disorders appears higher for COS than in AOS

contrast groups (74,75).

In genetic studies of complex disease there are currently two salient working

hypotheses, the common disease–common allele model, which proposes that

combinations of common genetic variations (i.e., polymorphisms in >1% of the

population) contribute a modest effect to disease cause or susceptibility (182).

Alternatively, the common disease–rare allele hypothesis proposes that some

individually rare mutations (<1% of the population, which a single individual or

family may have) with high penetrance contribute to disease cause or susceptibility

(183).

Common Variant Alleles

While now considered generally inadequate, earlier studies of common variants

focused on candidate genes, which utilized existing biologic knowledge to identify

single-nucleotide polymorphisms (SNPs) as markers to map and trace transmission

of regions of the chromosome (182). One of the key issues about this approach is that

many disease genes have been unanticipated based on what was previously known.

Notwithstanding, using the hypothesis that there may be more detrimental/penetrant

mutations in known schizophrenia susceptibility genes in COS, several genes

previously implicated in AOS samples showed significant association with COS

including G72 (now DAOA) (184), neuregulin 1 (NRG1) (185), GAD1 (186), and

dysbindin (DTNBP1) (187). In addition to supporting the genetic continuity between

COS and AOS, these data suggested that the very early-onset COS population, with

more pronounced neurobiologic abnormalities and a more homogeneous phenotype,

might turn out to be a relatively efficient population to target for the identification of

genetic risk of schizophrenia more generally.

With the technologic advances in microarray analyses, genome-wide association

studies (GWASs) have opened a new window into the genome at a submicroscopic

level by systematically assessing a broad array of SNPs independent of prior

scientific knowledge (188). Because GWASs require stringent corrections for

multiple comparisons, studies must have very large numbers of participants/samples

in order to be adequately powered. This encouraged international collaborative

efforts such as the Psychiatric Genomic Consortium (PGC) in order to pool samples.

Despite these efforts, samples for rare diseases such as COS remain small with

pooled numbers of less than 300 which cannot compare to AOS studies of over

16,000 patients (189). Results from large GWAS studies in AOS have included the

association of the major histocompatibility complex and the intron 4 transcription

factor (TCF4) involved in neurogenesis and brain development (190). GWAS results

have also supported polygenicity, with multiple small genetic variations conferring

risk for AOS, genetic overlap with other mental illnesses, and difficulties replicating

findings from prior candidate gene studies (191–193). To overcome the issue of

small sample size, the NIMH team created polygenic risk scores for 130 COS

participants and their healthy sibling controls, using 80 genomic variants associated

to schizophrenia PGC GWAS. Despite the small sample size, COS participants had

higher risk scores than their healthy siblings (p < 0.05). Although only with the most

liberal significance threshold, the study also described overlap between polygenic

risk for autisms and COS (194).

Rare Variant Alleles

Studies of rare variant alleles have focused primarily on structural variations in

stretches of DNA known as copy number variations (CNVs). Essential to these

studies is that these types of variations are common in human populations and, while

they may denote disease risk, they may also represent standard genetic variation. One

of the initial seminal studies of CNVs in schizophrenia found a progressive increase

in novel CNVs with earlier age of onset of schizophrenia. While only 5% of the

controls had any CNVs >100 kb, 15% of AOS cases and 20% of young onset (< age

18) had CNVs. An independent replication in the same study found that 28% of COS

patients had CNVs compared to 13% in their control sample of COS parents. A large

number of the CNVs affected genes in brain development and regulation pathways

(183). This study was followed by numerous additional findings of CNVs in

schizophrenia (195–200). Using whole genome sequencing, the NIMH team assessed

CNVs in 126 COS patients and 69 of their healthy full siblings. Focusing on a group

of 46 CNVs associated to risk for AOS, autism, intellectual disabilities, and seizures,

the study found that COS patients had not only significantly higher rates of diseaserelated

CNVs compared to their healthy sibling controls (p = 00.17), but also than

that reported in AOS (p < 0.0001) (201).

Cytogenetic Abnormalities

High-resolution banding karyotype and fluorescent in situ hybridization (FISH)

analyses are done routinely on all COS participants to look for fragile X, 22q11

deletions, Smith–Magenis (17p11.2del), and 15q11–q13 deletions/duplications.

Almost 10% of the NIMH COS participants show chromosomal abnormalities. The

22q11 deletion syndrome (22q11DS), manifest as velocardiofacial syndrome

(VCFS), is estimated to occur 1 in 4,000 live births (202), and is a known risk factor

for schizophrenia (203,204). Five out of 126 (4.0%) of the NIMH COS patients have

VCFS with spontaneous 22q11.2 deletion, a rate significantly higher than reported in

healthy controls (0.2%) (p < 0.0001) (205), AOS patients (0.3% to 1%) (206–208)

(p < 0.0001), or any clinical population at present. Despite the association between

autism and the 22q.11.2 deletion (209,210), and the presence of prepsychotic autism

spectrum disorders in 20% of our sample, none of the individuals with the 22q11.2

deletion had prepsychotic autism symptoms (201). Similarly, 7 out of 133 have large

chromosomal abnormalities such as Turner syndrome, XYY, trisomy X, or either

translocations or frameshift mutations (211,212).

TREATMENT STUDIES IN CHILDHOOD PSYCHOSES

Although rare, COS is a devastating disorder, frequently resistant to treatment, and

with an unfortunately narrow evidence base to guide treatment, particularly as there

are few trials comparing atypical antipsychotics, which have become the mainstay of

current treatment (213). Two prior randomized controlled trials established the

superiority of typical antipsychotics over placebo in COS (214,215). A single trial in

a small group of treatment refractory COS patients had demonstrated the efficacy of

clozapine over the typical antipsychotic haloperidol (216). However, as there was

no placebo arm in the study, it is hard to assess the true effect size for clozapine. A

later double-blind randomized controlled trial of comparing clozapine (n = 12) with

olanzapine (n = 13) showed a significant advantage for clozapine in the alleviation of

negative symptoms of schizophrenia, which was not correlated with improvement in

mood or extrapyramidal side effects. As anticipated, clozapine was associated with

more overall side effects, including enuresis, tachycardia, hypertension, and

significant weight gain by 2 years (217). The results of the NIMH cohort and studies

in AOS patients (218,219) show that clozapine has the greatest antipsychotic

efficacy, particularly in a pediatric population, with our recent study finding over

70% of the 120 children available at follow-up adhering to clozapine treatment for

more than 2 years, despite its side effects and need for close monitoring (220).

Adverse Effects of Clozapine

In spite of its unique efficacy for some COS patients, clozapine is associated with

several side effects, in particular agranulocytosis, weight gain, cardiovascular

changes such as postural hypotension and tachycardia, and incontinence. The NIMH

study has started addressing the questions of how to manage these side effects so that

these children can continue to stay on clozapine.

Neutropenia and Akathisia

Children and adolescents treated with clozapine have increased susceptibility to

neutropenia, particularly in male children with African American decent ( 221). This

can be successfully managed by addition of lithium (222). Similarly, akathisia, seen

only rarely in adults on clozapine, appears more common in children and can

frequently manifest as worsening of psychotic symptoms or agitation in children,

which frequently results in dosage increment. This side effect is responsive to

adjunctive propranolol treatment (223).

Weight Gain

Weight gain is a significant effect of atypical antipsychotics and appears more

pronounced in pediatric patients (224,225). Clozapine particularly has been noted to

cause significant weight gain during childhood (226). Although the mechanism of

weight gain in poorly understood, genetic risks (polymorphism in beta 3 and alpha-

1A adrenergic, 5HT-2C, TNF-alpha, and histamine receptors) and a number of

biochemical correlates (e.g., leptin, prolactin, triglyceride, and HDL levels) of

weight gain have been reported in the literature (227). In our group, an analysis of 23

COS patients treated with clozapine and 21 matched healthy controls showed

increases in BMI (p < 0.001) and leptin levels (p = 0.003) after 6 weeks of

treatment. For COS patients, BMI at baseline and week 6 correlated with insulin

level (r = 0.5, p = 0.004) and BMI was positively correlated with clinical

improvement in CGI, SAPS, and SANS rating scales (p < 0.05) (228). Based on

these correlations, we have used the antidiabetic medication metformin (which

improves peripheral insulin sensitivity) with some success although no formal trial

was done.

Treatments for schizophrenia and psychosis

Pharmacological

While the efficacy of antipsychotics is broadly similar in children,

young people, and adults, the young show a greater

sensitivity to a range of antipsychotic-related adverse events,

including extrapyramidal side effects (EPS), treatment resistance

with traditional antipsychotics (Kumra et al., 1998b), and

weight gain, obesity, and metabolic syndrome with the newer

atypical antipsychotics (Correll et al., 2009).The pharmacology

of antipsychotic drugs and management of adverse effects is

discussed in Chapter 43.

Although the number of RCTs of antipsychotics in children

and young people (<18 years) with schizophrenia remains

small, in recent years there have been trials reported for aripiprazole

(Findling et al., 2008), risperidone (Haas et al., 2009),

olanzapine (Kryzhanovskaya et al., 2009), and paliperidone

(Singh et al., 2011). Taken together, antipsychotics show a small

but significant effect (ES=0.3) in reducing psychotic symptoms

(NICE, 2013).While antipsychotics do not differ significantly in

terms of efficacy, differences in side effects are notable; weight

gain is greatest with olanzapine (with up to 8 kg gain reported

in the first 12 weeks of treatment), intermediate with clozapine,

quetiapine, and risperidone and least with aripiprazole (Correll

et al., 2009). EPS are greater with “typical” antipsychotics such

as haloperidol and higher dose (>4mg) risperidone than with

quetiapine or olanzapine. Aripiprazole may cause akathisia

at higher doses. Sedation is greater with olanzapine, clozapine,

quetiapine, and haloperidol than with risperidone (Toren

et al., 2004) and aripiprazole. Hyperprolactinaemia is greatest

with risperidone, amisulpride, and haloperidol and least with

quetiapine and aripiprazole (aripiprazole may lower prolactin

levels from baseline).

Clozapine has been shown in randomized double blind

head-to-head trials to be superior to haloperidol (Kumra et al., use is restricted under license in the United Kingdom to

treatment-resistant schizophrenia defined as follows: (i) nonresponse

with at least two antipsychotics (drawn from different

classes and at least one being an atypical) each used for at least

4–6 weeks, and/or (ii) significant adverse effects with conventional

antipsychotics.Adverse effects of clozapine includeweight

gain, sedation, hypersalivation, and reduced seizure threshold.

The risk of blood dyscrasias on clozapine is effectively managed

by mandatory routine blood monitoring.

Baseline investigations and monitoring

Before starting treatment with antipsychotic medication, a

physical examination should include height, weight (BMI, body

mass index), and cardiovascular system examination, including

pulse and blood pressure, and a neurological examination for

evidence of abnormal movements. Baseline laboratory investigations

include full blood count, liver function and electrolytes,

prolactin, fasting blood glucose, HbA1C, and plasma lipids.The

NICE (2013) guidance recommends that weight be measured

weekly for the first 6 weeks of treatment, again at 12 weeks

and thereafter 6-monthly and pulse, blood pressure, prolactin,

fasting blood glucose, HbA1C and lipids should be repeated at

12 weeks and thereafter 6-monthly.

Pharmacological interventions: summary

Choice of antipsychotic medication is determined primarily

by side effect profile, given broadly similar efficacy of antipsychotics.

The greater sensitivity of children and young people

to adverse effects of antipsychotics means that the benefit to

risk ratio may be lower than in adults. Drug choice should be

a collaborative exercise, tailored to the needs and preferences

of the young person and the family. In children and young

people naive to antipsychotics, atypical drugs are preferred

as they minimize the risk of acute dystonic reactions that

are common with high-potency conventional antipsychotics

(e.g., haloperidol) in young people. Antipsychotics should be

started below the usual dose range for adults and titrated slowly

upwards against clinical response and side effects to determine

theminimum effective dose. Short-termuse of benzodiazepines

(e.g., lorazepam) is preferable to high-dose antipsychotics in

the management of severe behavioral disturbance associated

in acute psychosis. The recommended order of treatment for

first-episode schizophrenia in children and adolescents is:

atypical as first line; if inadequate response, change to a different

atypical or conventional antipsychotic; if response is still

inadequate or side effects are intolerable, then initiate clozapine.

Antipsychotic medication is usually maintained for at least

2 years after a first episode of schizophrenia, with formal review

of medication, including physical health, conducted at least

annually (NICE, 2013).

Psychosocial intervention

Cognitive behavior therapy (CBT)

In contrast to the adult literature, there is a much smaller and

weaker evidence base to support CBT and family interventions

in EOS. The NICE Guideline for schizophrenia and psychosis

(NICE, 2013) identified only six RCTs (N=460) comparing

individual CBT with a control intervention in children and

young people with mean age 25 years and younger (Power et al.,

2003; Haddock et al., 2006;Mak et al., 2007; Jackson et al., 2008;

Jackson et al., 2009; Edwards et al., 2011).There were no studies

of CBT where all participants were <18 years. Three of the studies

(Power et al., 2003; Jackson et al., 2008; Edwards et al., 2011)

were conducted in a specialist EPPIC in Australia. All participants

in these studies received the relatively intensive “treatment

as usual” (TAU) offered by EPPIC. Taken together, these studies

found no benefit of CBT compared to control interventions for

psychotic symptoms (Jackson et al., 2008; Edwards et al., 2011),

depression/suicidality (Power et al., 2003; Edwards et al., 2011),

or social adjustment (Jackson et al., 2008).Haddock et al. (2006)

reported an interaction with age, so that for young people <age

21, results were better for supportive counselling compared to

CBT or wait-list control,while in contrast, for young people age

>21, outcomes were superior for CBT.

These results are intriguing as they suggest that for young

people (<21 years) CBT may be relatively less effective as

an intervention for schizophrenia than in adults. Reasons

for this apparent lack of benefit of CBT in EOS may include

(i) methodological weaknesses and small numbers of young

people included in these studies and (ii) the early phase of the

illness and primary outcomes chosen. The strongest evidence

for CBT in adults with schizophrenia is for treatment-resistant

positive symptoms (Turkington & Kingdon, 2000), while in

trials in young people, CBT has been evaluated in the acute

phase of the illness. (iii) Greater cognitive impairments and/or

illness severity in young people may moderate the effectiveness

of unmodified CBT interventions developed for adults with

schizophrenia.

Family intervention

The UK NICE Guideline for schizophrenia and psychosis in

children and young people identified only two RCTs (N=158)

comparing family intervention with an active control in young

people aged <25 years. Both studies were conducted with

young people in remission for psychosis and compared family

CBT with individual CBT (Linszen et al., 1996; Gleeson

et al., 2009). Family intervention was no more effective than

an active control in reducing the number of participants who

relapsed. In summary, family intervention in young people

with schizophrenia fails to show the clear benefits for relapse

prevention (Number Needed to Treat of 4; see Chapter 14)

reported in adults (NICE, 2013).