The NIMH Treatment Units for Research on Neurocognition in Schizophrenia Network implemented the 4-week, placebo-controlled, parallel group, double-blind study. Inpatients or outpatients aged 18 to 60, who met DSM-IV-TR criteria for schizophrenia, were selected for study entry. Participants were diagnosed based on information from the Structured Clinical Interview for DSM-IV (29
), direct assessment, family informants, and past medical records. Participants were required to be clinically stable, in the non-acute phase of their illness, and to meet the following inclusion criteria (30
): a) treatment with one second generation antipsychotic medication, other than clozapine, for the previous two months, with no dose change in the month prior to study entry; b) Brief Psychiatric Rating Scale (BPRS; 31) hallucinatory behavior and unusual thought content item scores ≤4; c) BPRS conceptual disorganization item ≤4; d) Simpson-Angus Scale (SAS; 32) total score ≤ 6; and e) Calgary Depression Scale (CDS; 33) total score ≤10 (30
). In order to facilitate recruitment, the above criteria were amended halfway through the study to allow treatment with no more than two second generation antipsychotic medications, other than clozapine, and the cut-off score for BPRS hallucinatory behavior and unusual thought content items was changed to ≤ 5.
Participants were required to validly complete the MATRICS Consensus Cognitive Battery (MCCB; 35,36), i.e., the neuropsychological tester and the site neuropsychologist judged their performance to reliably reflect their ability on those aspects of cognition that the test was intended to measure. To minimize potential ceiling effects, participants were required to score at least one standard deviation below maximum on one or more of the following tests: Letter-Number Span; Hopkins Verbal Learning Test, and Identical Pairs Continuous Performance Test (CPT). Finally, participants were required to have a Wechsler Test of Adult Reading (37
) raw score ≥ 6.
Participants were excluded if they had a DSM-IV diagnosis of alcohol or substance abuse (other than nicotine) within the last month, alcohol or substance dependence (other than nicotine) within the last 6 months, or mental retardation; had a history of significant head injury/trauma or clinically significant medical or neurological disease; were treated with drugs known to act at the GABAA receptor or to inhibit CYP3A4; had a history of severe benzodiazepine withdrawal; or participated in a clinical trial of investigational medication within 60 days. Women of childbearing age were included if using adequate birth control.
The institution IRBs approved the study protocol and informed consent procedures. Written informed consent was obtained from all participants after study procedures had been fully explained and prior to study participation. Participant ability to provide valid informed consent was documented using study specific procedures.
The MCCB was used to assess neuropsychological test performance. The MCCB is comprised of 10 tests, which assess seven cognitive domains (35
). The MCCB composite score is a standardized mean of the seven domain scores. T-scores are standardized to normative data, and have an estimated mean of 50 and SD of 10 in the general healthy population (36
In addition, because of their previous use to evaluate cognitive effects with this compound (28
), the AX-CPT (38
) and the N-Back (39
) were used to assess prefrontal cortical cognitive function. The AX-CPT is a modification of the traditional CPT, in which AX trial frequency is increased to 70%. The increased AX trial frequency requires greater use of context to overcome the induced propensity to respond to the “X” probe on trials that do not contain the “A” cue (40
). The N-Back is a sequential letter working memory task, which varies working memory load by requiring the participant to identify whether the test stimulus is identical to the immediately preceding letter (0-back), the letter presented 1 trial back (1-back) or two trials back (2-back) (39
A modified version of the UCSD Performance-Based Skills Assessment (UPSA; 40), the UPSA-2, was used to assess functional capacity. The UPSA-2 contains a sixth component: Medication Management, and the content complexity and number of items required to be remembered were increased for the Comprehension/Planning, Financial Skills, and Transportation components to reduce potential for ceiling effects. The Schizophrenia Cognition Rating Scale (SCoRS; 41) is an interview-based measure used to assess cognition. The MCCB, UPSA-2, SCoRS, AX-CPT and N-back were obtained at Evaluation Week 1 and Treatment Phase Week 4. The BPRS positive symptom item total score was used to assess positive symptom change.
The BPRS positive symptom items are: conceptual disorganization, hallucinatory behavior, unusual thought content, and suspiciousness. The modified Scale for the Assessment of Negative Symptoms (SANS; 42) total score was used to assess negative symptom change. The CDS was used to assess depressive symptom change. The CGI severity of illness item (CGI-S) was used to assess global changes. The BPRS, SANS, CDS and CGI-S were obtained at Screening, Evaluation Phase Week 2 and biweekly during the Treatment Phase.
MCCB and UPSA raters were trained on the administration and scoring of these instruments using video and group training sessions and were individually certified by an expert on these assessments. The SCoRS raters were trained in a group education format, in which they viewed and scored a series of videotapes. Symptom raters were required to be reliable on the BPRS and SANS (ICC ≥ 0.80). Quarterly reliability meetings were conducted throughout the study to ensure that the raters maintained the inter-site ICC criterion of ≥0.80. All raters were blind to treatment assignment.
The Simpson Angus Scale (SAS; 32) and Abnormal Involuntary Movement Scale (AIMS; 43) were used to assess abnormal motor movements. The SAS and AIMS were administered at Screening, Evaluation Phase Week 2 and biweekly during the Treatment Phase.
A standard chemistry panel, complete blood count, urinalysis and urine toxicology screen, and EKG were obtained at Screening and at the end of the Treatment Phase. The Side Effect Checklist (SEC) was used to assess side effects and monitor vital signs. The SEC is comprised of 22 common side effects, which are rated from 1 (none) to 4 (severe). The SEC and vital sign ratings were conducted at Evaluation Phase Weeks 1 and 2 and weekly during the Treatment Phase.
Participants who met inclusion criteria entered a 2-week Evaluation Phase during which they underwent baseline cognitive, symptom and safety assessments. Participants who continued to meet inclusion criteria entered the 4-week, double-blind Treatment Phase and were randomized to MK-0777 3mg BID; MK-0777 8mg BID; or placebo BID. Participants randomized to MK-0777 8mg BID were started on 3mg BID and their dose was titrated over the first week to the target dose. Participants randomized to MK-0777 3mg BID were started on this dose. The MK-0777 t1/2 is approximately 7 hours and the Tmax is 6–7 hours, therefore we used a twice daily dosing schedule.
If side effects interfered with the tolerability of the study medication, the participant was instructed to skip a dose and then resume treatment with the prescribed dose. If still unable to tolerate the study medication, then the dose could be lowered to alleviate side effects. The side effects most likely to affect MK-0777 tolerability were dizziness, incoordination, and sedation. At the end of the Treatment Phase, all participants were tapered off of their study medication to minimize potential withdrawal effects.
The study biostatistician established computer-generated randomization sequences for each site. Randomization was performed using the permuted block method, randomly drawing from 3 or 6 size blocks, in order to limit imbalance in numbers between groups. Until the trial was concluded, the randomization sequence was only available to the biostatistician and to an unblinded pharmacist at each site, whose only role was to dispense medication. In response to a randomization request, the biostatistician sent a code number to the unblinded pharmacist, which identified the next treatment selection to be dispensed from the treatment sequence. Randomization was stratified by site.
Medication compliance was assessed by weekly pill count. All participants who received 75% or more of their assigned study medication were considered compliant.
The sample size was determined using the analysis of covariance power formula, n=2[za
, with za
=0.842 (corresponding to power=0.80), R=the correlation between baseline and end of study measures of the primary outcome (estimated to equal 0.6 for the MCCB composite score), d the difference between groups, and s the standard deviation of the primary outcome. We planned to enroll 30 participants per group, which would have enabled us to detect an effect size=0.73 with power=0.80. The actual recruitment was only about 20 participants per group, but the observed R
0.9, suggesting power to detect an effect size of 0.49.
An analysis of covariance (ANCOVA), adjusting for baseline scores, was used to compare treatment groups on cognitive and functional measures. The predefined primary cognition outcome measure was the MCCB composite T-score, tested at overall two-sided alpha=0.05. The predefined primary functional outcome measure was the UPSA summary score. Exploratory analysis of variation in treatment effects among the different MCCB measures was performed using the mixed model for repeated measures ANCOVA: week 4 T-score=baseline T-score + measure + treatment + treatment × measure, where measure was a categorical variable indicating the different MCCB tests, and the treatment × measure effect tested whether the treatment effect differed significantly among the various tests.
AX-CPT and N-back accuracy results were summarized using the d-prime statistic (44
). For the AX-CPT, only BX trials were used to calculate the false alarm rate. For the N-back, trials with novel and repeated distractors were pooled in calculating the false alarm rate. N-back response times (RTs) were analyzed using the ANCOVA model log(RT)=baseline log(RT) + response type + treatment + treatment × response type, where response type distinguishes target, repeat non-target and novel non-target trials.
Symptom data were analyzed using a mixed model for unbalanced repeated measures ANCOVA, using data from all participants who completed at least one symptom assessment to fit the model: follow-up score=baseline score + treatment + week + treatment × week, where the treatment effect tests the average difference across weeks between treatment groups, and the treatment × week interaction assesses whether this difference varies between weeks 2 and 4. Mixed models were fitted with SAS PROC MIXED®, using the Kenward-Rogers method to estimate degrees of freedom. The treatment × week interaction was non-significant for all variables assessed, and only average difference tests and estimates are reported.
Group differences on SAS and AIMS total scores were examined by calculating the tau-b rank correlation between score and week for each participant, and comparing the distribution of these trend scores using the Conover-Salsburg rank test (45
). Fisher’s exact test was used to compare treatments on the number of participants who, at any point during follow-up, had new or worsened (compared to baseline) side effect severity. The effects of treatment on laboratory assays were tested using ANCOVA, while the effects of treatment on vital signs were tested using mixed model ANCOVA.