Alzheimer’s disease (AD) is a progressive neurodegenerative disease, involving a large number of genes, proteins and their complex interactions. As an irreversible progressive brain disease, it can slowly destroy memory and thinking skills, then make people ultimately lose the ability to perform the simplest tasks. It is estimated that the morbidity of AD over the age of 65 could reach up to 10–50%. Due to its high social and economic costs, AD is considered as one of the most intractable medical problems. To date, there are no effective drugs or treatments available to prevent or reverse the progression of the disease.
Despite extensive research on neurodegenerative diseases and drugs, its clinical progress remains elusive. For example, recently, several potentially disease-modifying agents have been suggested for AD. Many of these suggested therapeutic agents have passed the efficacy testing in animal models. However, all of the ensuing phase three clinical trials have failed. These failures question our accurate understanding of the disease. In addition, they also show us the difficulty in clinical trial design of therapeutic agents.
That’s why you need an experienced and professional team to help you with innovations of trial design and management.
Case Study
Patient enrollment is a key challenge in most clinical trials. And studies of drugs to treat AD require an especially nuanced approach to recruiting. For example, syndromic definitions are critical to drug development. A specific patient population appropriate for the trial must be identified before the rating scale can be used to quantify the severity of the symptoms at baseline and to assess changes in symptoms during the trial. Researchers already have definitions of psychosis of AD and depression of AD and use them to inform previous trials. However, as the most common and challenging symptoms of AD and other NDD, there is no consensus definition of agitation. Here we use a provisional consensus definition of agitation. This definition was developed through a transparent, inclusive, reiterative process by the International Psychogeriatic Association (IPA).
Another challenge is statistical analysis. We used SAS version 9.2 and R version 2.13.1 for statistical analyses. All p-values are two-sided and p < .05 was the threshold for statistical significance. No adjustments were made for multiple comparisons.
We assessed primary efficacy based on intention-to-treat (ITT) comparison of the difference in the NBRS-A scores at week 9 and comparison at week 9 for the mADCS-CGIC. Crude between-treatment difference at week 9 NBRS-A scores was assessed using a t-test. To adjusted differences, we used mixed effects regression models with a random intercept for patient, indicators for each visit, treatment by visit interactions, and covarying for baseline NBRS-A scores and baseline MMSE (due to baseline imbalance). Mixed effects regression was also applied to estimate the difference in the linear slope of NBRS-A scores over all study visits. All available visit data for the 186 participants were incorporated into the NBRS-A model. Other continuous scale scores were modeled in the same way. Besides, we also performed sensitivity analyses for the NBRS-A outcome. First, we used generalized estimating equations (GEE) model for mean visit scores with unstructured covariance structure for within-person longitudinal measurements. Then the robust standard errors for effect was estimated. The mADCS-CGIC ratings of change (“marked worsening” to “marked improvement” on a 7 point scale) at week 9 were compared between treatment groups including all participants with week 9 mADCS-CGIC data using proportional odds logistic regression. We performed sensitivity analyses for the mADCS-CGIC outcome by using multiple imputation to estimate missing week 9 data. The proportion of participants experiencing adverse events was compared between treatment groups using Fisher's exact test for small cells (unadjusted), or logistic regression, adjusting for baseline report of the same symptom if necessary due to baseline imbalance. Adherence was assessed by pill counts from returned medication bottles using the Wilcoxon rank sum test.
Using this definition and statistical strategy, we recruited appropriate patients and analyzed results in a trial of citalopram treatment for agitation in Alzheimer's disease.
Design:
A multicenter, randomized, placebo-controlled, double-blind and parallel group trial, including complete eligibility criteria, data collection schedule and detailed statistical analysis.
Participants:
From 8 academic centers, there are 186 patients who may have AD and clinically significant agitation (Figure 1). All recruited patients all had performance as follows:
- Patients meeting the criteria are cognitively impaired.
- They show such behavior for at least 2 weeks.
- They are experiencing subjective distress.
- All of these patients exhibit motor hyperactivity, physical aggression, or verbal aggression.
- Patients have sufficient symptoms to cause disability, which can’t just attribute to cognitive impairment.
- Everyone has agitation that is not entirely caused by another disorder or by environmental circumstances.
Figure 1. Participant flow
Length of Enrollment Period:
42 months.
Interventions:
Participants (n=186) were randomized to receive a psychosocial intervention plus either citalopram (n=94) or placebo (n=92) for 9 weeks. Depending on response and tolerability, the dose began at 10 mg/d with planned titration to 30 mg/d over 3 weeks.
Main Outcomes:
- Neurobehavioral Rating Scale, Agitation Subscale (NBRS-A) - Primary Outcome Measures
- Modified Alzheimer Disease Cooperative Study-Clinical Global Impression of Change (mADCS-CGIC) - Primary Outcome Measures
- Cohen-Mansfield Agitation Inventory (CMAI)
- Neuropsychiatric Inventory (NPI)
- Activities of Daily Living (ADLs)
- Caregiver Distress
- Cognitive Safety (MMSE)
- Adverse Events
Results:
Participants who received citalopram treatment showed significant improvement in both primary outcomes compared with placebo. The NBRS-A estimated treatment difference at week 9 (citalopram minus placebo) was -0.93 [95% CI, -1.80 to -0.06], p = 0.036. Results from the mADCS-CGIC showed 40% of citalopram participants having moderate or marked improvement from baseline compared with 26% of placebo recipients, with estimated treatment effect (odds ratio [OR] of being at or better than a given CGIC category) of 2.13 [95% CI, 1.23-3.69], p = 0.007. Participants who received citalopram showed significant improvement on the CMAI, total NPI, and caregiver distress scores but not on the NPI agitation subscale, ADLs, or in less use of rescue lorazepam. Worsening of cognition (-1.05 points [95% CI, -1.97 to -0.13], p = 0.026) and QT interval prolongation (18.1 ms [95% CI, 6.1-30.1], p = 0.004) were seen in the citalopram group (Table 1).
Table 1. Primary and secondary outcomes
se = standard error, CI = confidence interval.
*The score and treatment effect are the model-based estimates calculated using mixed effects regression models. The treatment effect is the difference of the scores at week 9 controlling for baseline score and MMSE. A negative number favors citalopram for NBRS, CMAI, NPI and GUG. A positive number favors citalopram for MMSE and ADCS-ADL. 90 participants in citalopram and 85 participants in placebo had at least one NBRS follow-up measurement. 86 participants in citalopram and 81 participants in placebo had NBRS data at week 9. 85 participants in citalopram and 79 participants in placebo had MMSE data at week 9. For CMAI, NPI and ADCS-ADL, 86 participants in citalopram and 83 participants in placebo had data at week 9.
**The treatment effect estimate is the odds ratio (calculated using logistic regression) of using rescue lorazepam for citalopram vs. placebo. A number less than one favors citalopram. 90 participants in citalopram and 86 participants in placebo had data on lorzaepam use for at least one follow-up visit and were included.
†The treatment effect estimate is the odds ratio (calculated using proportional odds logistic regression) of being at or better than a given ADCS-CGIC category for citalopram vs. placebo. A number greater than one favors citalopram. 86 participants in citalopram and 81 participants in placebo had data on the ADCS-CGIC at week 9.
If you are looking for any specific service, please feel free to contact us.
References:
1. Liu H, et al. (2014) ‘Advances in recent patent and clinical trial drug development for Alzheimer’s disease’, Pharmaceutical patent analyst, 2014, 3(4):429-447.
2. Porsteinsson, A. P. , et al. (2014) ‘Effect of citalopram on agitation in Alzheimer’s disease – The CitAD randomized controlled trial’, JAMA : The Journal of the American Medical Association, 311(7), 682–691.
3. Cummings, J., Zhong, K. (2015) ‘Trial design innovations: Clinical trials for treatment of neuropsychiatric symptoms in Alzheimer’s Disease’, Clinical Pharmacology and Therapeutics, 98(5), 483–485.