Efficacy of Cognitive Remediation in Patients With Schizophrenia or Schizoaffective Disorder Stabilized on Lurasidone

OVERVIEW & SPECIFIC AIMS Marked cognitive impairment underlies much of the social &
occupational dysfunction associated with schizophrenia. Currently available antipsychotic
medications are primarily effective in treating psychotic symptoms & have demonstrated only
limited potential in ameliorating cognitive deficits in schizophrenia patients.

Lurasidone is a novel compound synthesized by SEPRACOR, Inc.for the treatment of patients
with schizophrenia & bipolar disorder. It possesses high affinity for dopamine D2, serotonin
5-HT2A, 5-HT7, 5-HT1A & noradrenaline α2C receptors. Compared with other atypical
antipsychotics, lurasidone demonstrates similar binding affinities for the D2 & 5-HT2A
receptors, but greater affinity for serotonin 5-HT1A receptors. Lurasidone displays no
affinity for histamine H1 or acetylcholine M1 receptors. In animal studies, lurasidone
significantly reversed memory impairment induced by MK-801, an N-methyl-D-aspartate (NMDA)
receptor antagonist, in a rat step-through type passive avoidance task. The maximum
inhibitory effects of lurasidone were greater than those observed with risperidone,
quetiapine, & olanzapine, while aripiprazole was not effective in reversing the impairment
induced by MK-801. Additionally, lurasidone significantly reversed memory impairment induced
by the anticholinergic drug scopolamine in the passive avoidance task. The reversal of
pharmacologically induced cognitive deficits in rats by lurasidone is promising & warrants
specific investigation in subjects with schizophrenia, given the prominence of cognitive
deficits in this disorder.

From a different therapeutic perspective, the utility of cognitive remediation in
ameliorating cognitive deficits & improving functional outcomes in schizophrenia has
recently been evaluated in several studies. A meta-analysis of these trials found effect
sizes for improvement in cognitive & psychosocial functioning in the low to moderate range
(McGurck 2007). The best outcomes in psychosocial functioning were evident when cognitive
remediation was combined with teaching of psychosocial skills.

Given the recalcitrant nature of cognitive deficits in schizophrenia & their impact on
functional capacity we felt that in designing a study to test the effectiveness of cognitive
remediation we should maximize the likelihood of therapeutic benefit by administering
cognitive remediation in the context of pharmacotherapy that may have potential for
precognitive effects. By so doing we could possibly boost the effect sizes seen with
cognitive remediation alone. In this study we will transition patients with schizophrenia
(in whom a change in antipsychotic therapy is clinically warranted) from their current
antipsychotic to lurasidone - clinicians will have eight weeks to complete the switch.
Subjects who are successfully switched to lurasidone will then be randomized to receive
either cognitive remediation or a non-specific mental activity control condition two
times/week for a total of 30 sessions over a 4-6 month period. Our goal is to have 140
patients complete the cognitive remediation phase.

A subset of the sample will participate in 2 biomarker studies. Event related potentials &
fMRI will be done in these subjects at baseline & study completion.

This study will be done as an Investigator initiated trial (J. Lieberman, M.D. - PI) under a
separate IND.

Primary Aim: We hypothesize that cognitive remediation will be superior to the active
control group on the change from baseline to study end point of cognitive remediation phase
on both co-primary outcome measures (standardized composite MATRICS score & Cognitive
Assessment Interview).

Additional aims

1. To compare cognitive remediation to active control on functional outcome as assessed by
the change in UCSD Performance-Based Skills Assessment (UPSA-Brief) from baseline to
end point of cognitive remediation phase.

2. To compare cognitive remediation to active control on changes from lurasidone
stabilized baseline to end point in indices of functional brain activation (ERP & fMRI)
during cognitive activation tasks.

3. Evaluate the effect of 8 weeks of lurasidone treatment on cognitive & functional
outcomes as assessed by changes from baseline in the MATRICS composite score, CAI, &
UPSA-Brief.

4. Evaluate the effect of cognitive remediation compared to nonspecific mental activity on
cognitive & functional outcomes as assessed by changes from lurasidone stabilized
baseline to end of cognitive remediation phase in the MATRICS composite score, CAI, &
UPSA-Brief.

5. Evaluate the efficacy, safety, & tolerability of lurasidone in patients with
schizophrenia as assessed by the change from baseline to week 8 & to end of cognitive
remediation phase in the PANSS total score, Side Effect Checklist, AIMS, SAS, BAS, &
frequency of abnormal laboratory values.

BACKGROUND & SIGNIFICANCE

A broad range of neurocognitive abnormalities characterizes patients with
schizophrenia. These include impairments in attention (including abnormalities in
sensory gating), visual & verbal learning & memory, working memory, processing speed,
social cognition, & executive function, (Nuechterlein et al, 2004), & are major
determinants of poor functional outcome (Green, 1996; Green et al, 2004). Conventional
antipsychotics have limited effects on these impairments. Second generation
antipsychotics may have modest benefits for cognitive function although recent
literature does not support an advantage over first generation antipsychotics (Davidson
et al, 2009). Regardless, most patients continue to exhibit pronounced cognitive
impairment despite adequate antipsychotic treatment & these deficits result in much of
the social & occupational dysfunction associated with schizophrenia.

A number of investigations have examined the impact of cognitive remediation on
cognition in schizophrenia subjects. A recent meta-analysis of cognitive remediation
trials (McGurck et al. 2007) found a low to moderate effect size for improvement in
cognitive performance (effect size 0.41), & psychosocial functioning (effect size
0.36). Interestingly, the effects of cognitive remediation on psychosocial functioning
were significantly greater in those studies that provided adjunctive psychiatric
rehabilitation (effect size 0.47) than in those that provided cognitive remediation
alone (effect size 0.05). In other words, a basic cognition-enhancing strategy had to
be combined with the teaching of psychosocial skills & strategies to see clinically
meaningful effects in psychosocial functioning (which is the ultimate goal of any
cognitive remediation intervention).

Of all the cognitive deficits in schizophrenia, verbal learning & memory are among the
most abnormal. Abnormalities are present at the earliest stages of auditory processing
as evidenced by the abnormally low amplitudes of the mismatch negativity response
obtained during the pre-attentive detection of auditory stimuli. Reduced mismatch
negativity responses are significantly associated with impaired verbal memory, with the
inability to decode semantic & emotional aspects of speech, & with poor functional
status. These findings suggest that efficient auditory processing is crucial for the
successful encoding & retrieval of verbal information & that disturbance in these
elemental processes are related to higher-order cognitive dysfunction in schizophrenia
(Fisher et al 2009).

Based on this body of evidence & on data from animal experiments in the basic
neuroscience of learning-induced neuroplasticity, a cognitive training program that
targets both early auditory processing & working memory operations was developed by
Posit Science with the goal of improving verbal memory performance in patients with
schizophrenia. Interim findings from an ongoing study to evaluate the effectiveness of
this approach were recently published (Fisher et al 2009). 55 patients with
schizophrenia who were clinically stable on various antipsychotics were assigned either
to the cognitive training program developed by Posit Science (50 sessions administered
5X/week over 10 weeks), or to a computerized game control condition to mimic the time &
concentration at a computer required in the cognitive training program. Primary outcome
was the change from baseline in the Measurement & Treatment Research to Improve
Cognition in Schizophrenia (MATRICS) test battery. The group that received cognitive
training showed impressive & significantly larger improvements than the control group
on global cognition, verbal working memory, & verbal learning & memory at effect sizes
in the medium to large range (.56 to .86). Important questions not addressed by this
study were the optimal dose & duration of the cognitive remediation intervention, the
tolerability & efficacy of the intervention in below average IQ schizophrenia
populations, the persistence (or lack thereof) of the improvement in cognitive function
after cessation of the cognitive remediation, the functional significance of the gains
in cognitive function, & the contribution, if any, of the type of antipsychotic
therapy. Additionally, it is unlikely that patients in typical clinical settings would
be able to comply with such a rigorous daily cognitive remediation program without the
financial incentives that were provided to subjects in this study.

As noted above, studies of cognitive remediation have thus far included subjects on
whatever antipsychotic they were taking at study entry. It is well known that the
antihistaminergic & antimuscarinic properties of antipsychotics may further impair
cognitive processes in schizophrenia patients. Additionally, antipsychotics vary
considerably in their EPS liability, which can affect cognitive processing & motor
speed. Thus, choice of antipsychotic therapy would appear to be an important
consideration in the design of cognitive remediation trials in schizophrenia patients
although no studies thus far have attempted to control for this important variable.

Lurasidone is a novel compound synthesized by Sepracor for the treatment of patients
with schizophrenia & bipolar disorder. It possesses high affinities for dopamine D2,
serotonin 5-HT2A, 5-HT7, 5-HT1A & noradrenaline α2C receptors. Compared with other
atypical antipsychotics, lurasidone demonstrates similar binding affinities for the D2
& 5-HT2A receptors, but greater affinity for serotonin 5-HT1A receptors. Lurasidone
displays no affinity for histamine H1 or acetylcholine M1 receptors. As of September
02, 2008, approximately 1560 people had received lurasidone in various types of
clinical studies (e.g., single doses ranging from 0.1-100 mg, repeated doses up to 600
mg/day for less than one week, & repeated doses up to 120 mg/day for 6 weeks of
treatment & up to 80 mg/day for 12 months of treatment). This included approximately
263 healthy volunteers & 1309 schizophrenic or schizoaffective disorder subjects.
Another 600 subjects are currently enrolled in ongoing, placebo-, & active comparator
controlled Phase 3 studies. Double-blind placebo-controlled & open label studies
lasting 6-8 weeks in patients with acute exacerbation of schizophrenia, were suggestive
of a therapeutic dose range of lurasidone of 40mg/day to120 mg/day given once daily.

Of special interest in animal studies was the observation that lurasidone significantly
reversed memory impairment induced by MK-801, an N-methyl-D-aspartate (NMDA) receptor
antagonist, in a rat step-through type passive avoidance task. The maximum inhibitory
effects of lurasidone were greater than those of risperidone, quetiapine, & olanzapine,
while aripiprazole was not effective in reversing the impairment induced by MK-801.
Additionally, lurasidone significantly reversed memory impairment induced by the
anticholinergic drug scopolamine in the passive avoidance task. The reversal of
pharmacologically induced cognitive deficits in rats by lurasidone & the differences
observed from other antipsychotics is promising & warrants specific investigation in
subjects with schizophrenia.

In this study we propose to build on the important work done thus far in cognitive
remediation in schizophrenia by addressing some of the limitations of previous studies.
All subjects will be stabilized on lurasidone prior to initiation of cognitive
remediation. This will provide a uniform foundation of antipsychotic therapy with an
agent that might have pro-cognitive effects. Cognitive remediation initiated in
schizophrenia patients stabilized on lurasidone treatment could potentially provide a
synergistic benefit on cognition & functioning. At a minimum, lurasidone would not be
expected to worsen cognitive functioning, as it has no M1 or H1 antagonist activity.
The cognitive remediation intervention will be administered in small group settings
twice weekly for 30 sessions & will utilize computerized & verbal group training
exercises to address basic skills such as auditory processing, attention, processing
speed, & verbal working memory & learning, as well as intermediate & complex skills
such as deductive reasoning, planning & sequencing, set shifting, & complex problem
solving. Software exercises from Posit Science will be used to address the deficits in
basic cognitive skills. Cognitive remediation sessions will include a 55 minute session
in which the subject engages in cognitive exercises on the computer followed by a 20
minute "bridging" component in which participants gather in a circle to discuss skills
practiced, how they might use that skill in real world tasks, & to learn other
non-computer based techniques to enhance their cognitive & psychosocial skills in order
to meet their overall recovery goals. Outcome measures include improvement in cognition
as assessed by the MATRICS & CAI, & functional outcome using UPSA-B. Two biomarker
assessments (ERP & fMRI) will also be conducted in a subset of subjects at baseline &
study completion.

Our goal in this study is to conduct a scientifically rigorous clinical trial that
optimizes the likelihood of meaningful improvement in cognitive & psychosocial
functioning in patients with schizophrenia. An important consideration in the study
design is that the intervention is practical & can be implemented in regular clinical
settings. Biomarker assessments will provide valuable information on neuroanatomical &
pathophysiological substrates of cognitive impairment & the effect of therapeutic
interventions.

PHARMACOLOGIC TREATMENTS

The daily dose of lurasidone will be flexibly dosed at 40, 80 or 120 mg based solely on
clinical considerations & investigator judgment. All these doses have been demonstrated
in large double-blind placebo-controlled trials to be effective in the management of
acute exacerbation as well as maintenance treatment of schizophrenia.

COGNITIVE REMEDIATION

The Cognitive Remediation program has two principal parts: (1) individualized
computer-based cognitive training allows participants to hone their cognitive skills
through repeated exercises which gradually increase in difficulty & complexity (2)
"bridging" group therapy sessions facilitate the transfer & application of newly
learned cognitive skills to everyday situations & recovery goals. A building block
approach to session planning will be used, with learning blocks increasing in
complexity from Basic to Intermediate to Complex. Three dimensions determine complexity
of the tasks:

- Cognitive skill emphasized by the activity (one versus multiple skills)

- Cognitive load of the activity (difficulty of task)

- Goal properties of the tasks (proximal versus distal, highly specific versus
complex)

Within each learning block are sessions that introduce an assortment of computerized
cognitive exercises, which build upon previous skills learned, but are not reliant on
achieving 100% accuracy before progressing. Thus foundation skills need to be
completely mastered to progress onto more complicated tasks. The overarching training
model is to introduce basic cognitive skills in the beginning block, with tasks that
target one or two cognitive skills, have proximal goals, & low cognitive load. In
addition to exercising the requisite basic cognitive skills, this facilitates
self-efficacy & task valuation, which in turn enhances motivation. Increasingly tasks
become more complex & touch upon various cognitive domains. In this manner, multiple
domains are engaged simultaneously - as is the case when the participant must navigate
everyday life outside the laboratory setting.

The "building block" approach allows the participant to gradually train each cognitive
skill necessary for higher order executive functioning. The ultimate goal is to teach
memory techniques & the problem-solving skills to successfully steer through tasks not
only on the computer but in everyday life

Non-specific mental activity Active Control Group (ACG). Participants will be enrolled
in twice weekly group that works on various computerized puzzles & mentally stimulating
exercises for total of 30 hours within a 6 month timeframe. Commercially available
puzzles games that provide engaging but non-specific mental stimulation will be used as
the active control training exercises. Participants will follow similar session
timeframe as CR groups followed by healthy lifestyle groups.

Structured Clinical Interview for DSM-IV-Clinical Trial (SCID-CT) The Structured
Clinical Interview for DSM-IV Clinical Trial Version (SCID-CT) will be used to confirm
the diagnosis of schizophrenia. The SCID-CT is a semi-structured interview designed to
evaluate DSM-IV Axis I diagnoses. It enables trained clinical raters to reliably
determine Axis I diagnoses in diverse patient populations.

Clinician-Rated Side-Effect Scales Patients are to be assessed for extrapyramidal
symptoms such as bradykinesia, rigidity, tremor, hyperkinesias, dystonia, akathisia,
involuntary muscle contractions, athetosis, & chorea.

EPS-related side effects will be evaluated with the standardized scales of AIMS, SAS &
BAS.

Abnormal Involuntary Movement Scale (AIMS) The AIMS consists of 12 items, 10 to be
rated on a 4-point scale of severity & 2 to be rated as yes or no.

Simpson-Angus Scale (SAS) The SAS is a rating scale used to measure EPS & consists of a
list of 10 symptoms, each to be rated on a 5-point scale of severity.

Barnes-Akathisia Rating Scale (BARNES) The Barnes Akathisia Rating Scale will be used
to assess the presence & severity of akathisia before treatment (baseline) & at the
scheduled visits. This scale consists of 4 items (objective observation of akathisia by
the investigator, subjective feelings of restlessness by the patient, patient distress
due to akathisia, an global evaluation of akathisia). To complete this scale, subjects
should be observed while they are seated & then standing (for a minimum of two minutes
in each position). Symptoms observed in other situations (e.g., while engaging in
neutral conversation, engaging in activity) may also be rated. Subjective phenomena
should be elicited by direct questioning.

The Positive & Negative Syndrome Scale (PANSS) The Positive & Negative Syndrome Scale
(PANSS) will be the primary assessment instrument for psychopathology.

The PANSS contains 30 items that assess symptoms of psychotic disorders including
positive, negative & general psychopathology. The PANSS was chosen because of its
widespread use in clinical studies of psychosis, & its demonstrated reliability in
assessing psychopathology across diverse patient populations. The PANSS includes items
from the Brief Psychiatric Rating Scale (BPRS), with additional items from the
Psychopathology Rating Schedule.

Clinical Global Impressions Severity Scale (CGI-S) The Clinical Global Impressions
(CGI) Severity Scale will be used for repeated evaluations of global psychopathology.
The CGI-S scale is widely used in schizophrenia research & is a single Likert scale
rating severity of psychopathology on a scale of 1 (normal, not ill) to 7 (very
severely ill).

Calgary Depression Scale for Schizophrenia (CDSS) The CDSS is a nine-item scale
specifically developed for assessment of depression in patients with schizophrenia.
Compared with the HAM-D, there is less overlap between positive & negative
psychopathology. Items do not focus on weight change & initial insomnia, both of which
can be confounded by the drug treatment of schizophrenia. All items are rated on a
four-point scale: 0=absent; 1=mild; 2=moderate; 3=severe.

Columbia Suicide-Severity Rating Scale (CSSRS) The CSSRS is a low-burden,
clinician-administered tool that covers the wide spectrum of suicidality from ideation
to behavior. With suicide attempts too infrequent to serve as an outcome parameter,
this suicide assessment scale provides a validated measure of such related variables as
impulsivity, poor frustration tolerance, sadness, & hopelessness.

Service Utilization & Resources Form -Short Form (SURFs) The Service Use & Resources
Form for Schizophrenia (SURFs) is a multi-item self-report form that comprehensively
documents use of mental health & non-mental health service use, including the number &
duration of contacts & the specific types of agencies from which services were
obtained.

BIOMARKER STUDIES

A total of 72 subjects (36 per group) will participate in each biomarker study (ERP &
fMRI). However we still anticipate a 50% total attrition from baseline to end point &
will therefore have to conduct baseline biomarker assessments in 144 subjects to yield
72 completers & 58 subjects with good data.

EVENT RELATED POTENTIALS

Verbal episodic memory processing will be examined by measuring event-related brain
potentials (ERPs) during a continuous word recognition test, & behavioral measures of
verbal working memory will be obtained using a word serial position test (WSPT), both
of which we have extensively used in our studies of schizophrenia (Bruder et al., 2004;
Kayser et al., 1999; Kayser et al., 2006; Kayser et al., 2009). We have chosen to use
the auditory version of both tests because ERP deficits were found to be particularly
evident for spoken words (Kayser et al., 2009) & this will enable assessment of changes
in auditory processing following the cognitive remediation with Posit Science auditory
training. ERP & behavioral assessments will be conducted at baseline (on previous
antipsychotic prior to switching to lurasidone) & at study end point upon completion of
cognitive remediation.

ERPs During Verbal Episodic Memory: Studies using neuropsychological tests have found
that patients with schizophrenia display a selective deficit in verbal learning &
memory (Saykin et al., 1991). Neuroimaging studies have linked the verbal episodic
memory deficits in schizophrenia to left inferior frontal & medial temporal lobe
structures (Gur et al., 1994; Mozley et al., 1996). PET studies indicate that left
inferior frontal cortex is involved in both encoding & retrieval of words in a
recognition memory task, & schizophrenic patients show reduced activation of this
region during both stages of episodic memory (Ragl& et al., 2001).

Episodic memory processes have also been examined by measuring brain ERPs during a
continuous word recognition memory task (Friedman, 1990b). Subjects view or hear a
series of words, some of which are repeated after a number of intervening words, &
their task is to decide whether each word is new (not previously presented) or old
(previously presented). A robust, replicable finding in healthy adults has been greater
ERP positivity to "old" as compared to "new" words about 250 to 800 ms after word
onset, referred to as the 'old-new effect.' The old-new effect has been interpreted as
a neurophysiologic representation of retrieval processes involved in consciously
discriminating old from new words. Intracranial ERP recordings in & around medial
temporal lobe structures of epilepsy patients have revealed similar old-new effects,
suggesting that the hippocampus, parahippocampal gyrus or amygdala may contribute to
the scalp recorded effects (Elger et al., 1997).

We have used a continuous word recognition paradigm to assess neurophysiologic
correlates of episodic memory deficits in schizophrenic patients & healthy controls
(Kayser et al., 1999). In our most recent study (Kayser et al., 2009), schizophrenic
patients showed poorer accuracy of word recognition, which is consistent with findings
of impaired verbal memory in schizophrenia. Patients showed marked reductions the ERP
old-new effect over lateral-parietal sites. Patients also lacked the
left-greater-than-right hemispheric asymmetry typically observed for healthy adults.
Impairments of these electrophysiologic correlates of episodic memory were most robust
for auditory stimuli, which suggests a left-lateralized deficit in encoding &/or
retrieval of phonological information. Patients also showed reduced left-parietal P3 &
response-related mid-frontal negativity, likely associated with performance monitoring
& anterior cingulate cortex.

We propose to record ERPs using essentially the same auditory word recognition task as
in our prior study (Kayser et al., 2009). Words synthesized for a male voice (484 ms
median duration) are presented binaurally through headphones at a comfortable level of
about 70 dB SPL. A constant 2.5 s stimulus onset asynchrony & a fixation cross to
minimize eye movements are used. The words are selected from a dictionary of 925 nouns.
Words are arranged in two lists of 80 items, pseudo-randomly assigning items to the
lists with the constraint that item features (e.g., concreteness or imagery norms for
words) are balanced across lists. For each list, an item sequence is constructed so
that an equal number of words (n = 17) are repeated once following either a short lag
(8 intervening items) or a long lag (24 intervening items). Thus, each sequence has 34
words that repeat once, 17 at each lag, & 46 filler words that do not repeat, yielding
a total of 114 items per sequence. Items that are repeated are considered new items at
the first presentation, & old items at the second presentation, & these repeated items
would form the basis for the subsequent data analysis. In contrast, never-repeated
words are considered filler items & not included in the data analysis. Word
presentation order is pseudo-randomized within each sequence to yield an equal
distribution of short & long lags. Participants are instructed to respond to every word
as quickly & accurately as possible by pressing one of two buttons on a response pad to
indicate whether the word was "new" (never presented in the series) or "old"
(previously presented). Prior to the first block, practice trials are administered with
stimuli not used in the test to ensure that subjects understand the task & are
responding appropriately. Each session will consist of two blocks of 114 trials with a
5 min rest interval between blocks. The retest session will also consist of two blocks,
but with different words. Each session will take about one hour including electrode
preparation. ERP & behavioral data will be acquired at each site & sent to NYSPI for
processing & analyses.

Verbal Working Memory & Tone Discrimination Tests: There is now considerable evidence
that working memory is impaired in schizophrenia (Goldman-Rakic, 1991; Park & Holzman,
1992; Carter et al., 1998). We have found that schizophrenic patients, who performed as
well as healthy controls on a tone discrimination test of perception & attention,
showed poorer performance than controls on an auditory verbal WM test, i.e., the Word
Serial Position Test (WSPT), which requires short-term storage of the serial position
of a series of four words (Bruder et al., 2004; Wexler et al., 1998). Moreover, when
recording ERPs during this working memory test we found that patients having
schizophrenia had reduced amplitude of P3 source over left inferior parietotemporal
region (Kayser et al., 2006).

The WSPT & a tone discrimination screening test will be administered behaviorally after
the ERP measures & take about 30 min. In the WSPT (Wexler et al., 1998), each trial
begins with a series of four nouns spoken in a male voice followed after a delay
interval of 9 sec by one of these words. The subject's task is to remember the four
words in the order presented & to indicate the position of the probe word in the
sequence. There are a total of 36 trials. Schizophrenic patients show poorer
performance than healthy controls on this verbal working memory test (Bruder et al.,
2004; Wexler et al., 1998). Moreover, patients who perform poorly on the tone screening
test of auditory discrimination & attention have a particularly marked deficit on the
WSPT & episodic memory tests, & may benefit most from the auditory training approach to
cognitive remediation (Posit Science). The tone discrimination task is simply to judge
whether two pure tones separated by 100 ms are the same or different in pitch. There
are 10 practice trials & 60 test trials consisting of 30 trials where the tone pairs
are the same pitch & 30 trials where the tones are different, with 6 trials at each of
five frequency ratios. The auditory ERP, WSPT, & tone discrimination test may be of
particular value as predictors of improvement in cognitive function following auditory
training & also for monitoring changes during cognitive remediation.

fMRI OF WORKING MEMORY & EPISODIC MEMORY

Rationale: Working memory (WM), or the ability to hold information "on line" for short
periods of time & often (but not always) to manipulate that information (Baddeley
1992), is considered to be one of the most fundamental cognitive impairments in
schizophrenia. It is closely related to the construct of executive dysfunction.
Deficits in both working memory & executive processing have been related particularly
to dysfunction of the prefrontal cortex (PFC) (Goldman-Rakic 1999; Weinberger 2001).
The exemplar of the executive WM task, is the "N-back" task (Cohen 1994). The "N-Back"
task requires "continuous updating & order memory." It has been used to assess WM
function in schizophrenia in several neuroimaging studies & has shown sensitivity to
change as a function of pharmacological challenge in healthy controls (Mattay 2003) &
cognitive rehabilitation in individuals with schizophrenia (Wykes, 2002). Because of
the relative simplicity of the N-back paradigm, activation can be measured in a shorter
time period & with fewer complexities than the Sternberg type task, & is likely to
yield robust data in a "blocked" design format in a substantially shorter time period
than storage tasks.

Episodic (Declarative) Memory Assessment Rationale: Episodic (declarative) memory will
be assessed using intentional word encoding followed by a yes/no recognition test.
Tests of this type robustly activate a range of brain regions known to be involved in
the pathophysiology of schizophrenia, including inferior frontal cortex,
parahippocampus, thalamus, parietal cortex & hippocampus. Individuals with
schizophrenia are impaired on episodic memory tasks, & on tasks of this nature
specifically. The paradigm chosen for this task is straightforward, easy for almost all
individuals with schizophrenia to complete, relatively easy to implement in a multisite
study, & will allow for the assessment of both basic components of episodic encoding &
retrieval, as well as proactive interference effects in secondary analyses.

Inclusion Criteria:

- Male or female between 18-55 years of age who meet DSM-IV-TR criteria for
schizophrenia or schizoaffective disorder confirmed by the Structured Clinical
Interview for DSM-IV Clinical trial version (SCID-CT version). Duration of illness >
1 year. Outpatient status.

- Change in antipsychotic medication is clinically warranted as evidenced by

- persistent psychosis despite adequate dose and duration of antipsychotic, or *
inability to achieve therapeutic dose because of dose-limiting side effects,

- persistent side effects that either cause significant subjective distress or
significantly increase medical risks, such as substantial weight gain or
metabolic disturbances, or

- patient preference to switch and treating psychiatrist is in agreement.

- No behaviors suggesting potential danger to self or others over the 6 months prior to
participation.

- For the last 2 weeks of lurasidone stabilization phase, a score of 4 or less on PANSS
items of conceptual disorganization, hallucinations, suspiciousness and unusual
thought content items.

- At end of lurasidone stabilization phase, Simpson-Angus Scale total score <

- At end of lurasidone stabilization phase, Calgary Depression Scale total score <10.

- No acute medical problems; any chronic medical condition (e.g. hypertension)
consistently treated and stable during the 1 month prior to participation.

- Able to provide signed informed consent and to cooperate with all study procedures.

- Able to attend twice weekly sessions (each lasting approximately 75 minutes) for
cognitive remediation or active control sessions for the ~6 month duration of the
cognitive remediation phase of the study.

- Must meet the following cognitive performance criteria:

- Able to complete the baseline MATRICS validly at baseline as assessed by NP
tester.

- Raw score of 12 or greater on the WTAR (Wechsler Test of Adult Reading) at
screening.

- Women who can become pregnant must be using an adequate method of contraception to
avoid pregnancy throughout the study and for up to 4 weeks after the study in such a
manner that the risk of pregnancy is minimized. Acceptable methods include oral,
injectable or implanted contraceptives, intrauterine devices or barrier methods such
as condoms, diaphragm and spermicides. Women who can become pregnant must have a
negative urine pregnancy test at the Screening Visit. Women who can become pregnant
include anyone who has experienced menarche and who has not undergone successful
surgical sterilization (hysterectomy, bilateral tubal ligation or bilateral
oophorectomy), or is not postmenopausal (defined as amenorrhea 12 consecutive
months).

Exclusion Criteria:

- Documented history of learning disability.

- Hearing or visual impairment; not fluent in English.

- Current treatment with clozapine or history of treatment resistance as evidenced by
failure to improve (in the judgment of the investigator) with 2 or more adequate dose
antipsychotic trials of at least 6 weeks duration in preceding 1 year.

- Concomitant or anticipated treatment with potent CYP 3A4 inhibitor such a cimetidine,
cyclosporine, erythromycin or erythromycin-like drugs (e.g., azithromycin,
clarithromycin except short term acute treatment for 1 week or less), diltiazem,
itraconazole, ketoconazole or other systemic antifungal agents in the azole class,
nefazodone; or potent CYP3A4 inducer including: carbamazepine, modafinil,
Phenobarbital, phenytoin, rifampin, St. Johns Wort, and troglitazone.

- Current treatment with psychotropic agents known to affect cognition such as
amphetamines, topiramate.

- History of treatment with electroconvulsive therapy within the 6 months prior to
participation or expectation that patient may require ECT during the study.

- History of neurological or neuropsychiatric conditions (e.g. stroke, traumatic brain
injury, epilepsy, etc).

- Subjects with a history of clinically significant neurological, metabolic, hepatic,
renal, hematological, pulmonary, cardiovascular, gastrointestinal, and/or urological
disorders (e.g. unstable angina, decompensate congestive heart failure, CNS infection
or history of HIV seropositivity), which would pose a risk to the patient if they
were to participate in the study or that might confound the results of the study.
Active medical conditions that are minor or well controlled are not exclusionary if
they do not affect risk to the patient of the study results. For example, the
following are not exclusionary: a) stable and well-controlled hypertension; b) asthma
(no serious attacks in the past year); c) hypothyroidism (TSH within normal limits).

- A positive test for Hepatitis C antibody with concurrent evidence of impaired hepatic
function (increased AST or ALT greater than 2 times the upper limit of normal) or
positive tests for Hepatitis A antibody IgM fraction or Hepatitis B surface antigen,
irrespective of the AST or ALT values.

- History of alcohol or substance abuse or dependence during the 6 months prior to
participation.

- Participation in a clinical trial involving an investigational medication within 3
months prior to participation or 2 or more investigational drug trials in the
preceding 12 months.

- Pregnant women or women of child-bearing potential who are not using adequate birth
control.

- Woman who are breast feeding.

- Individuals who: a) received any cognitive remediation in the 6 months prior to study
entry or b)received more than 6 hours of cognitive remediation in the 12 months prior
to study entry or c) received more than 15 hours in the 24 months prior to study
entry. Cognitive remediation is defined as any behavioral intervention consisting of
training activities that aim to target impairments in cognitive domains of sensory
processing, attention, memory, processing speed, working memory, and executive
functioning.