Treatment to Enhance Cognition in Bipolar Disorder (TREC-BD)

The purpose of the present study is to evaluate a neuroplasticity-oriented, computer-based cognitive remediation treatment program in patients with bipolar disorder and its effects on cognitive deficits and community functioning compared to an active, computer-based control.

Cognitive dysfunction is increasingly recognized as a major feature of bipolar disorder (BD), present by illness onset, persistent into euthymia, and associated with functional outcome. Deficits are qualitatively similar to those seen in schizophrenia (SZ), and may be quantitatively similar in some patient groups, e.g. in patients with a history of psychosis. Despite strong associations between cognitive impairment and functional outcomes in BD, treatment for these symptoms at present is inadequate. Pharmacotherapies do little to address cognitive symptoms, and may even worsen them. Psychosocial cognitive remediation (CR) treatments have been developed to target these symptoms and their functional correlates, and have shown early promise in patients with SZ in improving both neurocognition and community functioning. However, despite the overlap of neurocognitive deficits between patients with SZ and BD, no studies to date have extended neuroscience-based CR to patients with BD. The present study aims to assess the efficacy of CR treatment in patients with BD with a history of psychosis using a 70-hour CR paradigm compared to a dose-matched computer-based control. It is hypothesized that patients in the CR group will exhibit improvements in cognitive and community functioning compared to controls, which will persist during a 6-month durability phase. Additionally, putative mechanisms of functional change will be examined, including mediator effects of cognitive and clinical change on community functioning. 130 patients with BD with a history of psychosis recruited from the Psychotic Disorders Programs at McLean Hospital will be randomized into either the CR or computer control group. CR will be administered using the BrainWorks program, neuroscience-based training programs that have shown early promise in patients with SZ. Participants will be assessed on measures of clinical, cognitive, and community functioning at baseline, following the 70-hour treatment or control phase, and again 6 months later. Participants may opt to participate in an functional magnetic resonance imaging (fMRI) study at pre- and post-treatment; resting state, task-based and functional connectivity, and diffusion tensor imaging data will be collected to evaluate preliminary evidence of neurobiological changes after training versus control. Additionally, participants may opt in to participation in two tasks of reward sensitivity including the Probabilistic Reward task and Richard's Delay Discounting. These tasks are administered pre- and post-treatment to evaluate the role of reward in treatment response as well as the potential for CR to modulate reward processing. This project is in keeping with the NIH's stated strategic priorities for improving mental health outcomes in patients and strengthening the impact of National Institute of Mental Health (NIMH)-supported research on public health, with specific recommendations for broad implementation of effective psychosocial interventions.

This arm will receive computer-based cognitive remediation treatment 3 times per week for 24 weeks, for a total of 70 hours of treatment

Group will receive 70 hours of computer time playing pre-selected computer games administered in a similar format as the Cognitive Remediation condition

13 programs targeting cognition in 4 separate domains: Auditory processing, visual processing, social cognition, and executive functioning. Games are imbedded in a format that is engaging and interactive. Animated characters serve as "directors" for each program, explaining the tasks in both verbal and written formats and providing feedback on each trial and overall after each activity. Users move systematically through the programs and can track their progress as the go. Each session includes activities from several different games to maintain interest and train a variety of skills; however, games are presented in the order of domains listed above (i.e. auditory, then visual, then social, and finally executive) to avoid stimulus interference during the training.

Sessions will involve generic computer games administered via the game interface "Sporcle." Sporcle is a game site that offers a collection of quiz-type activities available on line. The activities include typing, basic timed arithmetic (e.g. simple addition); picture identification (e.g. name the corporate logo; name the sitcom), and subject-based quizzes in areas like history (e.g. name the US presidents), geography (e.g. name the state capitals), and literature (e.g. book title fill-in-the-blank). An administrator can track the activity of subjects including when they logged in, which games they played and for how long, and what their accuracy was on each game. We will use a pre-developed game schedule that includes a mix of each type of game in each session, and ensures that subjects are playing the same games in the same order. This format was developed to mirror the treatment condition, as subjects are given a variety of specific games to play at each session.

The Measurement and Treatment Research to Improve Cognition in Schizophrenia (MATRICS) battery includes 10 tasks that are designed to measure seven key cognitive domains: processing speed, attention, working memory, verbal learning, visual learning, problem solving and social cognition. These scores are also combined to yield a cognitive Composite. All subtest, domain, and composite scores are reported in standardized T scores with a mean of 50 and a standard deviation of 10; higher scores reflect better performance. For example, a score of 60 on any subtest, domain, or the Composite would represent a score 1 standard deviation above than the mean. All standardized scores are computed by the MCCB scoring software included in the testing battery, and are normed by age and sex. Total administration time is 60-90 minutes.

within 1 week prior to initiating intervention; midpoint - on average 8 weeks after initiation; post-treatment - on average 24 weeks after initiation; after 6 months no active intervention

The YMRS is an interview style measure asking about hallmark symptoms of mania. Total score ranges from 0 to 60 where higher scores indicate more severe symptoms of mania. Total score ≤12 indicates remission (13-19=minimal symptoms; 20-25=mild mania, 26-37=moderate mania, 38-60=severe mania). Administration time = approximately 10 minutes

within 1 week prior to initiating intervention; midpoint - on average after 8 weeks of initiation; post-treatment - on average 24 weeks after initiation; after 6 months of no active study intervention

The MADRS is an interview-style rating scale to assess severity of symptoms of depression. The MADRS consists of 10 items scored 0-6. Total MADRS scores range from 0-60, with higher score indicates more severe depression. Typical clinical cutoff points are: 0 to 6 - normal/symptom absent; 7 to 19 - mild depression; 20 to 34 - moderate depression; 34 - severe depression. Administration time = 10 minutes

within 1 week prior to initiating intervention; midpoint - on average after 8 weeks of initiation; post-treatment - on average 24 weeks after initiation; after 6 months of no active study intervention

The PANSS is an interview-administered measure assessing positive and negative symptoms of psychosis, and general psychiatric symptoms. The PANSS consists of 30 total items scored 1-7 (least to most severe). Both Positive and Negative sub scales consist of 7 items each for a total possible score of 49 for each sub scale; the General sub scale consists of 16 items for a total possible score of 112. As the lowest possible score is 1, the lower bound of PANSS total score is 30. Higher scores reflect greater symptom severity. Based on the authors' original publication Kay and colleagues reported mean score in a sample of people with schizophrenia as follows: Positive scale = 18.20 Negative scale = 21.01 General psychopathology = 37.74 Administration = approximately 40 minutes

within 1 week prior to initiating intervention; midpoint - on average after 8 weeks of initiation; post-treatment - on average 24 weeks after initiation; after 6 months of no active study intervention

The MCAS is an interview-based assessment that measures functioning in psychiatric patients in multiple domains including social interest and effectiveness, independence in daily living, and instrumental role functioning. The present study uses an abbreviated version of the form consisting of 11 total items scored 1-5, with higher scores reflecting better community functioning. The abbreviated version (Lewandowski et al., 2013) was selected because it assesses community functioning independent of cognition or clinical symptoms, which would represent a confound in the present study. Total possible scores range from 11-55, with higher scores reflecting better community functioning.

within 1 week prior to initiating intervention; midpoint - on average after 8 weeks of initiation; post-treatment - on average 24 weeks after initiation; after 6 months of no active study intervention

The SOFAS is a 100-point scale similar to the Global Assessment of Functioning designed to evaluate social and occupational functioning not directly influenced by psychological symptom severity. Assessment is based on rater impression and includes a single assigned number. Scores may range from 0-100, with higher scores reflecting better functioning.

within 1 week prior to initiating intervention; midpoint - on average after 8 weeks of initiation; post-treatment - on average 24 weeks after initiation; after 6 months of no active study intervention

within 1 week prior to initiating intervention; post-treatment - on average 24 weeks after initiation

Inclusion Criteria: Clinical diagnosis of BD with psychosis Positive and Negative syndrome Scale (PANSS) < 75; PANSS Psychosis item scores = 3 or under; Young Mania Rating Scale (YMRS) = 6 or under Age between 18 and 50 Within 10 years of illness onset Legal and mental competency of the participant Exclusion Criteria: Age under 18 or over 50 PANSS >75; PANSS Psychosis item scores >3; YMRS > 6 Legal or mental incompetence (legal incompetence defined by any guardianship (including of person or treatment guardianship); mental incompetence defined by failure of the informed consent survey) Psychiatric inpatient status at time of enrollment Delirium secondary to medical illness Psychotic or mood disorder due to general medical or neurological illness History of head trauma History of seizure disorder or photo-sensitive seizures Use of anticholinergic medication, clozapine or olanzapine at baseline Rapid-cycling bipolar disorder Diagnosis of current substance abuse (past month) or substance dependence within the past year.

Lewandowski KE, Sperry SH, Cohen BM, Norris LA, Fitzmaurice GM, Ongur D, Keshavan MS. Treatment to Enhance Cognition in Bipolar Disorder (TREC-BD): Efficacy of a Randomized Controlled Trial of Cognitive Remediation Versus Active Control. J Clin Psychiatry. 2017 Nov/Dec;78(9):e1242-e1249. doi: 10.4088/JCP.17m11476.

Lewandowski KE, Sperry SH, Ongur D, Cohen BM, Norris LA, Keshavan MS. Cognitive remediation versus active computer control in bipolar disorder with psychosis: study protocol for a randomized controlled trial. Trials. 2016 Mar 12;17(1):136. doi: 10.1186/s13063-016-1275-7.