Effect of Idalopirdine as Adjunct to Cholinesterase Inhibitors on Change in Cognition in Patients With Alzheimer Disease: Three Randomized Clinical Trials

Key Points

Question  Does idalopirdine, a selective 5-hydroxytryptamine-6 receptor antagonist, improve cognitive change in patients with mild to moderate Alzheimer disease when added to cholinesterase inhibitors?

Findings  In 3 randomized clinical trials that included a total of 2525 patients with Alzheimer disease treated with cholinesterase inhibitors, the added use of idalopirdine compared with placebo did not decrease cognitive loss over 24 weeks.

Meaning  The findings do not support the use of idalopirdine for the treatment of Alzheimer disease.

Importance  New therapeutic approaches for Alzheimer disease (AD) are needed.

Objective  To assess whether idalopirdine, a selective 5-hydroxytryptamine-6 receptor antagonist, is effective for symptomatic treatment of mild to moderate AD.

Design, Setting, and Participants  Three randomized clinical trials that included 2525 patients aged 50 years or older with mild to moderate AD (study 1: n = 933 patients at 119 sites; study 2: n = 858 at 158 sites; and study 3: n = 734 at 126 sites). The 24-week studies were conducted from October 2013 to January 2017; final follow-up on January 12, 2017.

Interventions  Idalopirdine (10, 30, or 60 mg/d) or placebo added to cholinesterase inhibitor treatment (donepezil in studies 1 and 2; donepezil, rivastigmine, or galantamine in study 3).

Main Outcomes and Measures  Primary end point in all 3 studies: change in cognition total score (range, 0-70; a lower score indicates less impairment) from baseline to 24 weeks measured by the 11-item cognitive subscale of the Alzheimer’s Disease Assessment Scale (ADAS-Cog); key secondary end points: Alzheimer’s Disease Cooperative Study–Clinical Global Impression of Change Scale and 23-item Activities of Daily Living Inventory scores. Dose group efficacy required a significant benefit over placebo for the primary end point and 1 or more key secondary end points. Safety data and adverse event profiles were recorded.

Results  Among 2525 patients randomized in the 3 trials (mean age, 74 years; mean baseline ADAS-Cog total score, 26; between 62% and 65% of participants were women), 2254 (89%) completed the studies. In study 1, the mean change in ADAS-Cog total score between baseline and 24 weeks was 0.37 for the 60-mg dose of idalopirdine group, 0.61 for the 30-mg dose group, and 0.41 for the placebo group (adjusted mean difference vs placebo, 0.05 [95% CI, −0.88 to 0.98] for the 60-mg dose group and 0.33 [95% CI, −0.59 to 1.26] for the 30-mg dose group). In study 2, the mean change in ADAS-Cog total score between baseline and 24 weeks was 1.01 for the 30-mg dose of idalopirdine group, 0.53 for the 10-mg dose group, and 0.56 for the placebo group (adjusted mean difference vs placebo, 0.63 [95% CI, −0.38 to 1.65] for the 30-mg dose group; given the gated testing strategy and the null findings at the 30-mg dose, statistical comparison of the 10-mg dose was not performed). In study 3, the mean change in ADAS-Cog total score between baseline and 24 weeks was 0.38 for the 60-mg dose of idalopirdine group and 0.82 for the placebo group (adjusted mean difference vs placebo, −0.55 [95% CI, −1.45 to 0.36]). Treatment-emergent adverse events occurred in between 55.4% and 69.7% of participants in the idalopirdine groups vs between 56.7% and 61.4% of participants in the placebo groups.

Conclusions and Relevance  In patients with mild to moderate AD, the use of idalopirdine compared with placebo did not improve cognition over 24 weeks of treatment. These findings do not support the use of idalopirdine for the treatment of AD.

Trial Registration  clinicaltrials.gov Identifiers: NCT01955161, NCT02006641, and NCT02006654

Alzheimer disease, a complex disease involving multiple pathophysiological mechanisms, is increasing in prevalence and cost due to the aging population.¹ Therapies with neurotransmitter-based mechanisms have the potential to affect multiple clinical domains in Alzheimer disease.

Two classes of medications are approved by the US Food and Drug Administration for the treatment of Alzheimer disease: cholinesterase inhibitors (donepezil, rivastigmine, and galantamine) and the N-methyl-d-aspartate–receptor antagonist memantine. Higher doses of cholinesterase inhibitors may be needed in patients with advanced Alzheimer disease, but this strategy is limited by increased adverse events due to peripheral nervous system cholinergic effects (eg, nausea, vomiting, and diarrhea).^2,3 Amplifying central nervous system cholinergic function without inducing peripheral nervous system cholinergic symptoms may therefore provide greater clinical benefits.

Evidence supports investigation of 5-hydroxytryptamine-6 (5-HT₆) antagonism in Alzheimer disease, which may involve modulation of cholinergic, monoaminergic, and glutamatergic systems.^4-6 The 5-HT₆ receptors affect learning and memory⁷ and the 5-HT₆ receptor antagonists have been shown to improve cognitive performance in animal models.^5,6 Even though clinical trials did not support 5-HT₆ antagonist monotherapy among patients with Alzheimer disease,^8,9 2 phase 2 trials suggested that administration of 5-HT₆ antagonists added to cholinesterase inhibitor therapy may improve cognition in Alzheimer disease.^10,11

Quiz Ref IDIn one of the phase 2 studies,¹⁰ 90 mg/d of idalopirdine (30 mg taken 3 times per day) added to a stable dose of donepezil provided significant improvement in cognitive performance relative to donepezil monotherapy among patients with moderate Alzheimer disease dementia. Therefore, 3 randomized clinical trials (STARSHINE [study 1], STARBEAM [study 2], and STARBRIGHT [study 3]) were conducted to assess the efficacy of idalopirdine when added to donepezil or another cholinesterase inhibitor therapy for the treatment of mild to moderate Alzheimer disease.

A phase 3 development program consisting of 3 randomized parallel-group, double-blind, fixed-dose, placebo-controlled 24-week studies was conducted. The phase 3 program also included a 28-week, open-label extension study (NCT02079246) for patients completing study 1 or study 2. The designs of the 3 studies are summarized in eFigure 1 in Supplement 1. All 3 studies were conducted in accordance with the principles of good clinical practice¹² and the Declaration of Helsinki.¹³

The trial protocol and the statistical analysis plan for study 1 appear in Supplement 2 and Supplement 3; study 2, Supplement 4 and Supplement 5; and study 3, Supplement 6 and Supplement 7. The local ethics committees approved all aspects of the trial design. Eligible patients or their legal representatives provided written informed consent before participating.

The 3 studies included patients (1) aged 50 years or older meeting the National Institute of Neurological and Communicative Disorders and Stroke–Alzheimer Disease and Related Disorders Association criteria¹⁴ for the diagnosis of probable Alzheimer disease, (2) with a Mini-Mental State Examination (MMSE) score at or between 12 and 22 at screening (range, 0-30; a lower score indicates higher impairment), and (3) taking a therapeutic and stable dose of a cholinesterase inhibitor for 4 months prior to screening. Patients were excluded if (1) taking memantine, (2) had an alternative cause of dementia, (3) had serious central nervous system or somatic disorders, (4) had clinically significant abnormalities (determined by laboratory testing), or (5) taking concomitant medications that would interfere with the safety and efficacy assessments.

The 3 studies used a parallel-group, fixed-dose design to explore the dose-response relationship for idalopirdine among patients receiving stable treatment with donepezil in studies 1 and 2 and any cholinesterase inhibitor (donepezil, rivastigmine, or galantamine) in study 3. Patents received 10 mg, 30 mg, or 60 mg of idalopirdine or an identical-appearing placebo (Figure 1, Figure 2, and Figure 3). Randomization and blinding were applied via an interactive voice response system to minimize risk of bias in the evaluation of the clinical effects of idalopirdine.

In all 3 studies, symmetric randomization to the groups was stratified by MMSE score stratum. In study 3, randomization was also stratified for base therapy (cholinesterase inhibitor therapy). Block randomization with block sizes of 3 (for study 1 and study 2) and 4 (for study 3) was used and restricted such that the first 2 patients at each site would always get different treatments. In study 3, Latin squares were used to balance treatments, strata, and sites. At least 50% of the patients were to be enrolled from the 12 to 18 MMSE score stratum (further details appear in the eMethods section in Supplement 1). Compliance was assessed via pill count at each visit.

During screening, patients were monitored by Lundbeck medical staff to assess eligibility prior to randomization. External quality oversight methods (including central review of scale administration) were used to try to achieve consistent and accurate ratings throughout the study for the 11-item cognitive subscale of the Alzheimer’s Disease Assessment Scale (ADAS-Cog),¹⁵ the 23-item Alzheimer’s Disease Cooperative Study Activities of Daily Living Inventory (ADCS-ADL₂₃),¹⁶ the Alzheimer’s Disease Cooperative Study–Clinical Global Impression of Change Scale (ADCS-CGIC),¹⁷ and the MMSE.

All study staff (including raters who performed the ADAS-Cog, ADCS-ADL₂₃, ADCS-CGIC, and MMSE assessments) were blind to patient randomization (ie, drug group assignment). The ADCS-CGIC rater was blind to any study assessment after baseline.

Quiz Ref IDThe primary and key secondary end points were identical across the 3 studies. The primary end point was change in cognition total score measured by the 11-item ADAS-Cog (range, 0-70; a lower score indicates less impairment)¹⁵ from baseline to 24 weeks. Key secondary end points were changes from baseline to 24 weeks in activities of daily living (function) measured by the ADCS-ADL₂₃ (score range, 0-78; a higher score indicates less impairment) and overall clinical response (global outcome) measured by the ADCS-CGIC.^16,17

At the baseline visit, the ADCS-CGIC was used to assess the severity of illness and has a rating from 1 (normal, not at all ill) to 7 (among the most extremely ill patients). At subsequent visits, the ADCS-CGIC was used to assess global change from baseline and has a rating from 1 (very much improved) to 7 (very much worse), with a rating of 4 indicating no change. These end points were assessed at baseline and at weeks 4, 12, and 24. The eMethods section in Supplement 1 contains a list of other prespecified end points. A safety follow-up visit occurred at week 28. For patients who withdrew from the studies, efforts were made to collect selected efficacy assessments after study drug discontinuation.

Safety data were collected via measurement of vital signs and weight, through physical and neurological examination, with clinical safety laboratory tests, by using electrocardiographic parameters, and via assessment of suicidality (using the Columbia-Suicide Severity Rating Scale).¹⁸ Qualified personnel coded adverse events using the lowest level term according to the Medical Dictionary for Regulatory Activities (version 19.0). According to the law of Hy, a potentially severe liver injury was defined as an alanine aminotransferase level or an aspartate aminotransferase level greater than 3 times the upper limit of normal in combination with a bilirubin level greater than 2 times the upper limit of normal. All criteria were measured as maximal values observed during the 3 studies; the criteria did not need to occur at the same time.¹⁹

Quiz Ref IDIn all 3 studies, the criterion for showing efficacy of a dose required demonstration of a significant positive effect on the ADAS-Cog followed by (in a gated procedure) a significant positive effect on either the ADCS-ADL₂₃ or the ADCS-CGIC. A significant effect meant that the adjusted mean difference for the idalopirdine dose vs placebo was statistically significant at a significance level of .05 using 2-sided testing. Due to the multiple end points and multiple active dose groups (10 mg, 30 mg, and 60 mg), multiple testing procedures were used to control the overall type I error in each study (details appear in the eMethods section in Supplement 1). For the testing strategies used to determine the positive or negative outcome of the trials, the direction of the effects on the end points was considered to ensure that a significantly deleterious treatment effect would not contribute to a positive outcome. All reported P values are 2-sided.

Based on the testing strategies used to control for multiplicity in the individual studies, sample sizes were determined to achieve 90% power for studies 1 and 3 and 80% power for study 2. For the power calculations, the assumed treatment differences and correlation of the end points were based on the outcome of a phase 2 study.¹⁰ The assumed difference in ADAS-Cog total score of −2 points was similar to the effects produced by cholinesterase inhibitors of −2.02 to −2.73 points in the midrange of the ADAS-Cog.²⁰

In each study, efficacy analyses were performed on the full analysis set, which was defined as all randomized patients who took at least 1 dose of investigational medicinal product and who had a valid baseline assessment and at least 1 valid postbaseline assessment of the primary end point. Changes from baseline in the ADAS-Cog, ADCS-ADL₂₃, and ADCS-CGIC were analyzed using a restricted maximum likelihood–based mixed model for repeated-measures approach. The model included terms to adjust for the effects of the MMSE strata and, in study 3 only, cholinesterase inhibitor therapy in addition to the effects of country and baseline scores. The effect of the idalopirdine doses was estimated as a mean difference vs placebo at week 24 using least-squares means for the treatment × visit interaction effect in the mixed model.

Sensitivity analyses that adjusted for the study site effects instead of country effects were done to evaluate the potential effects of site on the outcomes. Sensitivity analyses using a pattern-mixture model were done to evaluate the suitability of the missing at random assumption imposed by the mixed model. In these analyses, monotone missing values in patients withdrawn from treatment were imputed using multiple imputation based on the placebo group.²¹

The statistical analyses were performed using SAS version 9.4 (SAS Institute Inc). Additional details appear in the eMethods section in Supplement 1.

Among 2525 patients (mean age, 74 years; mean baseline ADAS-Cog total score, 26; between 62% and 65% of participants were women), 2254 (89%) completed the studies. The 3 studies were conducted in 34 countries worldwide from October 2013 to January 2017 (study 1: 933 patients at 119 sites in 16 countries; study 2: 858 patients at 158 sites in 18 countries; study 3: 734 patients at 126 sites in 16 countries). The final follow-up date was January 12, 2017.

Patient demographic and baseline clinical characteristics are presented in Table 1, Table 2, and Table 3. Patients had been diagnosed with Alzheimer disease a median of 1.4 to 1.9 years prior to enrollment, and had a mean duration of background cholinesterase inhibitor treatment of between 0.9 to 1.2 years. Patients had a mean MMSE score at baseline of approximately 18. The percentage of patients screening positive as carriers for at least 1 APOE ε4 allele ranged from 55.6% to 60.2% (study 1: 59.2%; study 2: 60.2%; study 3: 55.6%). The baseline patient demographics and clinical characteristics were not different between the idalopirdine and placebo groups across studies.

The patient withdrawal patterns were similar among the treatment groups; the proportion of total withdrawals was 9.7% for the 10-mg dose of idalopirdine, 9.7% for the 30-mg dose, 11.3% for the 60-mg dose, and 9.3% for placebo. In the full analysis sets, the proportion of missing postbaseline ADAS-Cog assessments (due to withdrawal, missed visits, or invalid assessment) was 3.7% in study 1 (Figure 1), 3.8% in study 2 (Figure 2), and 4.7% in study 3 (Figure 3).

At week 24, there were no significant differences in the change in cognitive performance as measured by the ADAS-Cog total score in the idalopirdine groups compared with placebo (Table 4 and Figure 4); therefore, Quiz Ref IDnone of the doses of idalopirdine met the criterion for efficacy in any of the studies.

In study 1, the mean baseline ADAS-Cog total score was 26.3 for the 60-mg dose of idalopirdine group, 26.7 for the 30-mg dose group, and 25.8 for the placebo group. The mean change in ADAS-Cog total score between baseline and 24 weeks was 0.37 for the 60-mg dose of idalopirdine group, 0.61 for the 30-mg dose group, and 0.41 for the placebo group. Compared with placebo, the adjusted mean difference in the change in ADAS-Cog total score between baseline and 24 weeks was 0.05 (95% CI, −0.88 to 0.98) for the 60-mg dose of idalopirdine group and 0.33 (95% CI, −0.59 to 1.26) for the 30-mg dose group.

In study 2, the mean baseline ADAS-Cog total score was 25.3 for the 30-mg dose of idalopirdine group, 25.9 for the 10-mg dose group, and 25.7 for the placebo group. The mean change in ADAS-Cog total score between baseline and 24 weeks was 1.01 for the 30-mg dose of idalopirdine group, 0.53 for the 10-mg dose group, and 0.56 for the placebo group. Compared with placebo, the adjusted mean difference in the change in ADAS-Cog total score between baseline and 24 weeks was 0.63 (95% CI, −0.38 to 1.65) for the 30-mg dose of idalopirdine group and −0.09 (95% CI, −1.10 to 0.92) for the 10-mg dose group. Given the gated testing strategy for statistical significance and the null findings at the 30-mg dose of idalopirdine, statistical comparison of the 10-mg dose vs placebo was not performed.

In study 3, the mean baseline ADAS-Cog total score was 26.2 for the 60-mg dose of idalopirdine group and 25.9 for the placebo group. The mean change in ADAS-Cog total score between baseline and 24 weeks was 0.38 for the 60-mg dose of idalopirdine group and 0.82 for the placebo group. Compared with placebo, the adjusted mean difference in the change in ADAS-Cog total score between baseline and 24 weeks was −0.55 (95% CI, −1.45 to 0.36) for the 60-mg dose of idalopirdine group.

As a consequence of the prespecified gated testing strategy for statistical significance and the nonsignificant findings for the primary end point, findings for the key secondary end points were considered nonsignificant de facto; therefore, no P values are reported for the key secondary end points. Similarly, the nonsignificant effects with the 30-mg dose of idalopirdine in study 2 implied nonsignificant effects with the 10-mg dose in that study.

The baseline and final values for the key secondary end points appear in Figure 4 and the within-group changes from baseline are reported in eTable 1 in Supplement 1. Individual patient-level data for the primary and key secondary end points are presented in eFigure 2 in Supplement 1. In all treatment groups and across all 3 studies, mean treatment adherence was greater than 98% and overall completion rates were high (89%-91%).

Adverse event data from studies 1, 2, and 3 are presented in Table 5, Table 6, and Table 7, including treatment-emergent adverse events reported by 3% or greater of the patients in any treatment group, the events leading to study withdrawal, serious adverse events, and cumulative increased levels of liver enzymes.

Quiz Ref IDThe most common treatment-emergent adverse event across treatment groups was accidental overdose (idalopirdine treatment groups: 5.2%-11.0%; placebo: 8.8%-11.8%) followed by falls (idalopirdine treatment groups: 4.2%-6.1%; placebo: 2.8%-6.0%). Cholinergic treatment-emergent adverse events, including vomiting and nausea, were noted at a numerically higher rate in the idalopirdine treatment groups (vomiting: 1.8%-4.1%; nausea: 1.6%-4.4%) compared with the placebo groups (vomiting: 0.7%-1.3%; nausea: 1.1%-2.5%). The proportion of study withdrawals due to treatment-emergent adverse events was 4.2% for the 10-mg dose of idalopirdine group, 5.1%-5.7% for the 30-mg dose groups, 5.8%-7.1% for the 60-mg dose groups, and 3.6%-5.3% for the placebo groups.

The incidence of an elevated alanine aminotransferase level or an aspartate aminotransferase level greater than 3 times the upper limit of normal were higher in the 3 idalopirdine dose groups (0.7%-3.3%) compared with the placebo groups (0%-1.3%) across studies. No patient fulfilled the criteria of the law of Hy (indicative of hepatic injury).¹⁹

The frequency of serious adverse events was low across system organ classes. A numerically higher percentage of patients in the 60-mg dose of idalopirdine groups (6.5%-7.7%) and in the 30-mg dose groups (5.3%-5.8%) experienced serious adverse events compared with the 10-mg dose group (4.6%) and the placebo groups (3.9%-6.0%). No specific category of serious adverse events explained the overall difference and an association with idalopirdine treatment was not established.

A total of 3 deaths occurred among patients receiving placebo (n = 955) and 5 deaths occurred among those receiving idalopirdine (2 in the 30-mg dose groups [n = 594] and 3 in the 60-mg dose groups [n = 672]) across the studies. Examination of reported causes of death revealed no indication of a treatment-related pattern. There were no notable differences in suicidal ideation or suicidal behavior among the treatment groups in any study.

The study results for the primary outcome remained unchanged in the sensitivity analyses using multiple imputation for patients withdrawn from the study based on values from the placebo group (eTable 2 in Supplement 1). Sensitivity analyses adjusting for site yielded results consistent with the main analysis that adjusted for country.

In these 3 randomized double-blind, placebo-controlled trials conducted in 2525 patients with mild to moderate Alzheimer disease in 34 countries, 6 months of idalopirdine treatment added to background cholinesterase inhibitor therapy did not improve cognition or mitigate the decline in symptoms measured by performance on the ADAS-Cog, ADCS-ADL₂₃, or ADCS-CGIC.

The hierarchical statistical testing strategy²² for each study dictated that (regardless of observations on key secondary outcomes, which measured domains other than cognition) failure to meet significance on the primary outcome equated to a null study. The findings for the key secondary outcomes measuring daily function (ADCS-ADL₂₃) and global clinical outcome (ADCS-CGIC) appeared congruent with a null effect. The results on 3 different dementia domains provide strong support for the absence of efficacy.

This study showed no efficacy in contrast to a phase 2 study¹⁰ that supported the potential efficacy of adding idalopirdine to a cholinesterase inhibitor. There were several major differences between the phase 2 study and the phase 3 program. These differences were in (1) dose (90 mg/d in phase 2 vs 10, 30, and 60 mg/d in phase 3) and schedule (30 mg taken 3 times per day in phase 2 vs 10, 30, and 60 mg/d in phase 3), (2) sample population (the phase 3 program broadened the range of cognitive impairment [MMSE score stratum of 12-22 vs 12-19 in the phase 2 study]), (3) scope of the phase 3 program, which was a multinational program conducted at hundreds of sites, and (4) allowable type of background cholinesterase inhibitor use (galantamine and rivastigmine also were allowed in study 3).

A different schedule and lower daily dose of idalopirdine were used in the phase 3 program compared with the phase 2 study. The results of an in vivo positron emission tomography receptor occupancy study conducted among healthy participants taking idalopirdine, which became available after completion of the phase 2 study, were used to guide dose and schedule selection in the phase 3 studies, as suggested by the US Food and Drug Administration.²³ A high 5-HT₆ receptor occupancy (>80%) was achieved at doses of 30 and 60 mg/d of idalopirdine, which was maintained for 24 hours, supporting the use of once daily dosing.²⁴ Although there is an approximate 40% decrease in 5-HT₆ receptor expression among patients with Alzheimer disease compared with healthy individuals, there are also potential limitations in translating occupancy data from healthy individuals to those with a diseased brain.²⁵

These 3 studies have several strengths. Each study tested multiple doses; used appropriate instruments for stage and purpose to assess multiple dementia domains and adverse events; had standardized training and monitoring of study site staff; used central monitoring of data and assessment methods; and had high completion rates. The demographics of the study patients were also representative of Alzheimer disease populations (eg, approximate mean age of 74 years; 65% were women; and 55%-60% were APOE ε4 carriers). Several characteristics support the external validity of this program, including the geographic, ethnic, and cultural diversity of the patients; inclusion of a broad range of cognitive scores (MMSE scores of 12-22); and relatively liberal inclusion criteria (eg, age ≥50 years with no upper limit, no central reading of brain magnetic resonance imaging or computed tomography, no requirement for Alzheimer disease profile positivity with amyloid positron emission tomography or cerebrospinal fluid profile). This phase 3 program also allowed concomitant use of any cholinesterase inhibitor (not just donepezil) as background treatment in study 3.

Placing the results of this study in the broader context of the 5-HT₆ antagonism adjunctive to cholinesterase inhibitor therapy mechanism of action suggests a lack of efficacy for this approach in the treatment of Alzheimer disease. In the phase 3 MINDSET study,²⁶ intepirdine (another 5-HT₆ antagonist) was administered at a dose of 35 mg daily (added to background donepezil therapy) in patients with mild to moderate Alzheimer disease and it failed to meet cognitive and functional efficacy outcomes. The negative results from the phase 3 MINDSET study run counter to the phase 2 results with intepirdine that had suggested a potential benefit in Alzheimer disease.¹¹

The 3 studies have limitations. There was no requirement for evidence of Alzheimer disease biomarker positivity for inclusion, which may have allowed some patients to be included without having Alzheimer disease pathology. In addition, background therapy with memantine was not allowed, which may limit extrapolation of the results to patients with Alzheimer disease with similar MMSE scores (ie, MMSE scores of 12-22) who are taking background monotherapy with memantine, or, more commonly, who are taking background combination therapy with a cholinesterase inhibitor and memantine.

In patients with mild to moderate Alzheimer disease, the use of idalopirdine compared with placebo did not improve cognition over 24 weeks of treatment. These findings do not support the use of idalopirdine for the treatment of Alzheimer disease.

Corresponding Author: Alireza Atri, MD, PhD, Ray Dolby Brain Health Center, California Pacific Medical Center Davies Campus, 45 Castro St, Ste 220, San Francisco, CA 94114 (atria@cpmcri.org).

Accepted for Publication: December 5, 2017.

Author Contributions: Dr Atri had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis. Drs Atri, Frölich, Ballard, Tariot, Molinuevo, and Cummings served as global, European, or US coordinating principal investigators for the idalopirdine phase 3 STAR clinical trial program.

Concept and design: Atri, Frölich, Ballard, Tariot, Boneva, Windfeld, Cummings.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: Atri, Frölich, Ballard, Molinueovo, Windfeld, Cummings.

Critical revision of the manuscript for important intellectual content: All authors.

Statistical analysis: Atri, Windfeld, Raket.

Administrative, technical, or material support: Raket, Cummings.

Supervision: All authors.

Conflict of Interest Disclosures: The authors have completed and submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Dr Atri reported receiving honoraria for consulting, providing educational lectures, programs, and materials, or serving on advisory boards for Allergan, the Alzheimer’s Association, Axovant, Biogen, Grifols, Harvard Medical School Graduate Continuing Education, Lundbeck, Merck, Sunovion, and Suven; receiving book royalties from Oxford University Press; and having institutional contracts or receiving investigational clinical trial–related funding from the American College of Radiology, AbbVie, Avid, Biogen, Lilly, Lundbeck, Merck, and vTV. Dr Frölich reported receiving honoraria for consulting, providing educational lectures, materials, and programs, or serving on advisory boards for Avid–Eli Lilly & Co, Avraham Pharmaceuticals, Axon Neuroscience, Boehringer Ingelheim, GE Healthcare, H. Lundbeck A/S, Merck Sharpe & Dohme, Novartis, Nutricia, Pfizer, Piramal Imaging, Schwabe Pharma, TAD Pharma, and Takeda. Dr Ballard reported receiving research grants from Acadia and H. Lundbeck A/S; and receiving honoraria from Acadia, Bristol-Myers Squibb, Heptares, Lilly, H. Lundbeck A/S, Novartis, Orion, Otsuka, and Roche. Dr Tariot reported receiving consulting fees from Abbott Laboratories, AbbVie, AC Immune, Auspex, Boehringer Ingelheim, California Pacific Medical Center, Clintara, CME Inc, Corium, GliaCure, INSYS, and T3D; receiving consulting fees and research support from AstraZeneca, Avanir, Cognoptix, Lilly, H. Lundbeck A/S, Merck and Company, and Takeda; receiving research support from Elan, Functional Neuromodulation, Genentech, Novartis, Roche, Targacept, the National Institute on Aging, and the Arizona Department of Health Services; owning stock options in Adamas; and being listed as a contributor to a patent owned by the University of Rochester. Dr Molinuevo reported receiving honoraria for consulting, providing educational lectures, or serving on advisory boards for ABL, Axovant, Boehringer Ingelheim, Eli Lilly & Co, Fujirebio, GE Healthcare, H. Lundbeck A/S, Merck Sharpe & Dohme, Novartis, Pfizer, Piramal Imaging, Roche, and Roche Diagnostics. Drs Boneva and Raket are full-time employees of H. Lundbeck A/S. Dr Windfeld was a full-time employee of H. Lundbeck A/S during the design and implementation of the studies reported herein and through the drafting and submission of the first version of the manuscript. Dr Cummings reported serving as a consultant to AbbVie, Acadia, Actinogen, Adamas, Alzheon, Anavex, Astellas, Avanir, Avid, Axovant, Boehringer Ingelheim, Bracket, Eisai, GE Healthcare, Genentech, Intra-Cellular Therapies, Lilly, H. Lundbeck A/S, MedAvante, Merck, Neurocog, Novartis, Orion, Otsuka, Pfizer, Piramal, QR Pharma, reMYND, Resverlogix, Roche, Suven, Takeda, Toyama, and Transition; receiving research support from Avid and Teva; owning stock options in Prana, Neurokos, Adamas, MedAvante, and QR Pharma; and owning the copyright of the Neuropsychiatric Inventory and all its derivatives. No other disclosures were reported.

Funding/Support: This research was supported by funding from H. Lundbeck A/S.

Role of the Funder/Sponsor: H. Lundbeck A/S was responsible for the design and conduct of the study; for the collection, management, analysis of data; and aided in the interpretation of the data. A medical writer employed by the study sponsor assisted in the preparation of the manuscript. The study sponsor had a role in the review and approval of the manuscript or the decision to submit for publication because 3 of the co-authors were employed by the sponsor at the time of the study.

Meeting Presentation: Presented in part at the Alzheimer’s Association International Conference; July 19, 2017; London, England.

Additional Contributions: We acknowledge the contributions of the principal investigators in these original studies; the Idalopirdine Development Teams; the STAR Program Teams, operational partners, and site staff and investigators; and those individuals serving on the data and safety monitoring board from the respective trials. In addition, we express our gratitude for the commitment of the study participants and their caregivers, without whose generous contributions and dedication this research would not be possible. Brian Odlaug, PhD (H. Lundbeck A/S), provided medical writing support in accordance with good publication practice guidelines.