Latrepirdine for Alzheimer's disease
Abstract
Background
Current treatments for Alzheimer’s disease (AD) provide modest symptomatic relief but do not slow the progression of the disease. Latrepirdine may modulate several targets involved in AD pathology, including lipid peroxidation, mitochondrial permeability, voltage‐gated calcium ion channels as well as neurotransmitter receptor activity, and thus potentially represents both a symptomatic and disease‐modifying intervention. Several randomized, placebo‐controlled trials have sought to evaluate the effect of latrepirdine on cognition, function and behaviour in patients with AD.
Objectives
To evaluate the efficacy and safety of latrepirdine for the treatment of AD.
Search methods
We searched the Specialized Register of the Cochrane Dementia and Cognitive Improvement Group on 4 June 2014 using the terms: latrepirdine OR dimebon OR dimebolin OR 2,3,4,5‐tetrahydro‐2,8‐dimethyl‐5‐ (2‐(6‐methyl‐3‐pyridyl)ethyl)‐1H‐pyrido(4,3‐b)indole.
Selection criteria
We included all randomized, double‐blind, placebo‐controlled trials where latrepirdine was administered to patients with mild, moderate or severe AD.
Data collection and analysis
We assessed the quality of studies and two authors extracted data. We calculated mean difference (MD), risk ratio (RR) and 95% confidence interval (CI) on an intention‐to‐treat (ITT) basis for all relevant outcome measures.
Main results
Seven trials involving a total of 1697 participants were found and six were included in the quantitative analyses. No data were available from the seventh trial. Three trials involving 1243 patients were included in analyses of efficacy outcomes, and four trials involving 1034 patients were included in analyses of safety and tolerability outcomes. We judged five trials to be at high risk of bias due to selective outcome reporting and three to be at high risk of attrition bias. There was low quality evidence favouring latrepirdine on the Clinician's Interview ‐ Based Impression of Change Plus Caregiver Input after 26 weeks (CIBIC‐Plus) (MD ‐0.60, 95% CI ‐0.89 to ‐0.31, 1 study, P < 0.001). Due to imprecision in the results, it was not possible to determine whether latrepirdine had any effect on cognition measured with the Alzheimer’s Disease Assessment Scale cognitive subscale (ADAS‐Cog) (MD ‐1.49, 95% CI ‐3.47 to 0.49, 3 studies, P = 0.14) or the Mini‐Mental State Examination (MMSE) (MD 0.59, 95% CI ‐0.94 to 2.11, 3 studies, P = 0.45), or on function measured with the Alzheimer’s Disease Co‐operative Study ‐ Activities of Daily Living scale (ADCS‐ADL) (MD 1.00, 95% CI ‐1.15 to 3.15, 3 studies, P = 0.36) at study endpoint (26 or 52 weeks). We considered the evidence provided on these outcomes to be of overall low quality. However, there was some high quality evidence showing a very small benefit of latrepirdine on the Neuropsychiatric Inventory (NPI) (MD ‐1.77, 95% CI ‐3.09 to ‐0.45, 3 studies, P = 0.009) at study endpoint (26 or 52 weeks). Additionally, moderate quality evidence suggested that latrepirdine and placebo were comparable in adverse events (RR 1.03, 95% CI 0.93 to 1.14, P = 0.51), serious adverse events (RR 0.86, 95% CI 0.55 to 1.35, P = 0.52), dropouts (RR 0.91, 95% CI 0.65 to 1.27, P = 0.57) and dropouts due to adverse events (RR 0.98, 95% CI 0.57 to 1.67, P = 0.93).
Authors' conclusions
Our meta‐analysis is limited by the small number of studies, imprecision, inconsistencies between studies and likelihood of bias. Nevertheless, the evidence to date suggests that while not associated with an increased risk of adverse events compared with placebo, there is no effect of latrepirdine on cognition and function in mild‐to‐moderate AD patients, though there appears to be a modest benefit for behaviour. Further studies should investigate the potential benefit of latrepirdine on neuropsychiatric symptoms in AD.
PICOs
Plain language summary
Dimebon for Alzheimer's disease
Alzheimer’s disease (AD) is a degenerative disorder of the brain. It is expected to be four times more common in 2050 than it was in 2006. Common symptoms of AD include memory loss, difficulties completing daily tasks, and changes in mood, behaviour and personality. Currently approved drug treatments for AD have only modest benefits and have not been shown to slow down the progression of the disease. There is a need for more effective drugs. Latrepirdine (also known as Dimebon) has been suggested as a potential treatment for AD and other dementias. Clinical studies have had conflicting findings, so we conducted a systematic review and pooled all the available data to assess the effects of latrepirdine. We looked for studies to help us answer this question by conducting a literature search in June 2014. We combined data from seven studies with a total of 1697 patients with AD. We were unable to conclude whether latrepirdine has any beneficial effect on cognition and function in people with AD due to variations in the results between studies and because the effects we estimated were too imprecise. However, the evidence suggests that latrepirdine may have a positive effect in treating behavioural symptoms, and that it is not associated with adverse effects in people with AD. Although seven studies have been done, data from only six studies were available and most of the studies were not fully reported. We contacted the investigators for additional data but received no response.
Authors' conclusions
Summary of findings
| Latrepirdine compared with placebo in Alzheimer's disease | ||||||
| Patient or population: patients with Alzheimer's disease | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Control | Latrepirdine | |||||
| CIBIC‐Plus (change from baseline at 26 weeks) | The mean CIBIC‐Plus (change from baseline at 26 weeks) in the control groups was 4.30 | The mean CIBIC‐Plus (change from baseline at 26 weeks) in the intervention groups was | 183 | ⊕⊕⊝⊝ | ||
| ADAS‐Cog | The mean change from baseline in ADAS‐Cog ranged across control groups from | The mean change from baseline in ADAS‐Cog in the intervention groups was 1.49 lower (95% CI ‐3.47 to 0.49) | 1243 | ⊕⊕⊝⊝ | 3 to 4 point change suggested to be clinically significant | |
| MMSE | The mean change from baseline in MMSE ranged across control groups from ‐1.90 to 1.19 | The mean change from baseline in MMSE in the intervention groups was 0.59 higher (95% CI ‐0.94 to 2.11) | 1243 | ⊕⊕⊝⊝ | 2 to 4 point change suggested to be clinically significant | |
| ADCS‐ADL | The mean change from baseline in ADCS‐ADL ranged across control groups from ‐6.91 to ‐0.74 | The mean change from baseline in ADCS‐ADL in the intervention groups was 1.00 higher (95% CI ‐1.15 to 3.15) | 1243 | ⊕⊕⊝⊝ | Limited data on clinically meaningful changes | |
| NPI | The mean change from baseline in NPI ranged across control groups from ‐1.04 to 3.65 | The mean change from baseline in NPI in the intervention groups was 1.77 lower (95% CI 3.09 to ‐0.45) | 1243 | ⊕⊕⊕⊕ | 4‐point change suggested to be clinically significant | |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 significant imprecision 3 significant heterogeneity across studies | ||||||
| Latrepirdine compared with placebo in measures of safety and tolerability in Alzheimer's disease | ||||||
| Patient or population: patients with Alzheimer's disease | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Control | Latrepirdine | |||||
| Adverse events | Study population | RR 1.03 | 1034 | ⊕⊕⊕⊝ | ||
| 583 per 1000 | 601 per 1000 | |||||
| Moderate | ||||||
| 569 per 1000 | 586 per 1000 | |||||
| Serious adverse events | Study population | RR 0.86 | 1034 | ⊕⊕⊕⊝ | ||
| 74 per 1000 | 63 per 1000 | |||||
| Moderate | ||||||
| 73 per 1000 | 63 per 1000 | |||||
| Total dropouts | Study population | RR 0.91 | 1034 | ⊕⊕⊕⊝ | ||
| 122 per 1000 | 111 per 1000 | |||||
| Moderate | ||||||
| 115 per 1000 | 105 per 1000 | |||||
| Dropouts due to adverse events | Study population | RR 0.98 | 1010 | ⊕⊕⊕⊝ | ||
| 51 per 1000 | 50 per 1000 | |||||
| Moderate | ||||||
| 46 per 1000 | 45 per 1000 | |||||
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 publication bias | ||||||
Background
Description of the condition
Alzheimer’s disease (AD) is a devastating and progressive neurodegenerative disorder with a global prevalence estimated at 26 million in 2006, a number expected to quadruple by 2050 (Brookmeyer 2007). Common symptoms include significant memory loss, difficulty completing daily tasks, a decline in problem‐solving ability and changes in mood, behaviour and personality (Alzheimer's Association 2010). There are currently many theories attempting to explain the pathobiology of AD. One major theory posits a dysfunction of the central cholinergic transmission and a loss of cholinergic neurons (Bassil 2009). The formation of toxic amyloid‐β (Aβ) deposits (Hardy 2002), neurotoxic aggregates of the tau protein (Anderton 2001), mitochondrial dysfunction (Bachurin 2003) and neuro‐inflammation (McGreer 2007) are several of the other well‐known hypotheses for AD pathogenesis. Despite massive global efforts to develop a successful treatment, disease‐modifying agents have yet to pass the clinical trial phases required for licensing approval. Four agents are currently available for the treatment of AD. These are donepezil, galantamine, rivastigmine (cholinesterase inhibitors (ChEI)) and memantine (an N‐methyl‐D‐aspartate (NMDA) antagonist) (Mangialasche 2010). These medications provide symptomatic relief but with modest effect sizes (Birks 2006; McShane 2006). Furthermore, they do not prevent or reverse the progression of the disease (Rafii 2009). Thus, there is an urgent need for more effective symptomatic and disease‐modifying interventions worldwide.
Description of the intervention
Latrepirdine (sold as Dimebon), an orally available non‐selective antihistamine developed in Russia and available since 1983, has recently been suggested as a possible treatment of AD and related dementias (Bachurin 2001).
How the intervention might work
At this time, the underlying mechanisms of action of latrepirdine are poorly understood (Wu 2008). Pharmacologic effects reported thus far include weak inhibition of butyrylcholinesterase, acetylcholinesterase and the NMDA receptor signalling pathway (Bachurin 2001). Unlike some other antihistamines (for example diphenhydramine), which are highly anticholinergic and have potentially negative cognitive effects (Kanamaru 2007), dimebon has been shown in vitro to weakly inhibit acetylcholinesterase and butyrylcholinesterase, thus potentially possessing pro‐cholinergic activity. In addition, latrepirdine has been shown to prevent lipid peroxidation and inhibit opening of the mitochondrial permeability transition pore as well as voltage‐gated calcium ion (Ca2+) channels in neurons (Bachurin 2003; Wu 2008). Importantly, while in vitro work showed protective actions against Aβ‐induced neurotoxicity, an in vivo study showed no effect on markers of Aβ‐induced neurotoxicity (Wang 2011).
Why it is important to do this review
Latrepirdine showed clinical benefits in an eight‐week, open‐label pilot study with 14 AD patients in Russia (Bachurin 2001). A subsequent phase II trial of latrepirdine, published in the Lancet in 2008, reported significant improvements in cognitive, functional, behavioural and global outcome measures over a six‐month period in a sample of 155 patients with mild‐to‐moderate AD (Doody 2008). However, the Global Phase III, Double‐Blind, Placebo‐Controlled Safety and Efficacy Study of Oral Dimebon in Patients With Mild‐to‐Moderate Alzheimer's Disease (CONNECTION) trial, a six‐month phase III confirmatory trial launched in 2008, demonstrated no significant effects of latrepirdine as a monotherapy compared with placebo in its co‐primary and secondary outcomes (Jones 2010). Another study, the Phase 3 Multicenter, Randomized, Placebo‐Controlled, Double‐Blind Twelve‐Month Safety and Efficacy Study Evaluating Dimebon in Patients With Mild‐to‐Moderate Alzheimer's Disease on Donepezil (CONCERT) trial, also did not meet the primary efficacy endpoints on cognition and function. Given these contradictory findings, a systematic review is required to determine the clinical efficacy and safety of latrepirdine in patients with AD.
Objectives
To evaluate the efficacy and safety of latrepirdine for the treatment of AD.
Methods
Criteria for considering studies for this review
Types of studies
We considered all randomized, controlled and double‐blind safety and efficacy trials of latrepirdine for patients with AD. We also included randomized controlled trials (RCTs) where latrepirdine given in combination with another anti‐dementia treatment was compared to the anti‐dementia treatment alone.
Types of participants
We included patients with a clinical diagnosis of AD with mild, moderate or severe dementia as determined by standardized diagnostic criteria, such as the Diagnostic and Statistical Manual of Mental Disorders IV, IV Text Revision and 5 (DSM‐IV, DSM‐IV‐TR, DSM‐5) (American Psychiatric Association 2013) and the National Institute of Neurological and Communicative Disorders and Stroke‐Alzheimer’s Disease and Related Disorders Association (NINCDS‐ADRDA) (McKhann 1984) criteria.
Types of interventions
Latrepirdine at any dosage compared with a placebo control arm.
Types of outcome measures
Primary outcomes
-
Clinical global impression of change (Clinician's Interview ‐ Based Impression of Change Plus Caregiver Input (CIBIC‐Plus), Alzheimer's Disease Cooperative Study‐Clinical Global Impression of Change (ADCS‐CGIC))
-
Cognitive function (Alzheimer’s Disease Assessment Scale cognitive subscale (ADAS‐Cog), Mini‐Mental State Examination (MMSE), Severe Impairment Battery (SIB))
Secondary outcomes
-
Behavioural symptoms including agitation and night‐time motor activity (Neuropsychiatric Inventory (NPI))
-
Mood including neurovegetative symptoms (e.g. sleep, appetite)
-
Functional performance or activities of daily living (Alzheimer’s Disease Co‐operative Study ‐ Activities of Daily Living scale (ADCS‐ADL))
-
Changes in global disease severity
-
Quality of life (EuroQol‐5D (EQ‐5D))
-
Caregiver burden and caregiver quality of life
-
Safety and tolerability
-
Dropouts, and dropouts due to adverse events
-
Institutionalization
-
Dependency
-
Mortality
-
Direct and indirect costs of health and social care
Search methods for identification of studies
Electronic searches
We searched the comprehensive register of dementia studies (ALOIS) created by the Cochrane Dementia and Cognitive Improvement Group on 4 June 2014 using the terms: latrepirdine OR dimebon OR dimebolin OR 2,3,4,5‐tetrahydro‐2,8‐dimethyl‐5‐(2‐(6‐methyl‐3‐pyridyl)ethyl)‐1H‐pyrido(4,3‐b)indole. This register contains reports of trials from bibliographic databases such as MEDLINE, EMBASE, PsycINFO and Web of Science, as well as references from searches performed in many trials registers and grey literature sources. This register is updated regularly and search strategies used for the retrieval of studies can be seen on the Group’s module in the Cochrane Library. We performed additional database searches in OvidSP (which includes references from MEDLINE, EMBASE, PsycINFO and EMB Reviews) on 4 June 2014 to ensure that as many relevant and up‐to‐date reports of trials were retrieved as possible. The search strategy that was created for the retrieval of references to studies in MEDLINE is available in Table 1. We also searched online trials registries (ClinicalTrials.gov, European Union (EU) Clinical Trials Register, World Health Organization (WHO) International Clnical Trials Registry Platform), conference proceedings, correspondence and news releases.
| Source | Search Strategy |
| Medline (Ovid SP) | 1. latriperdine.mp. 2. dimebon.mp. 3. dimebolin.mp. 4. 2,3,4,5‐tetrahydro‐2,8‐dimethyl‐5‐(2‐(6‐methyl‐3‐pyridyl)ethyl)‐1H‐pyrido(4,3‐b)indole.mp. 5. (1 OR 2 OR 3 OR 4) 6. (Alzheimer* OR AD).mp. 7. Alzheimer Disease/ 8. Dement*.mp. 9. (6 OR 7 OR 8) 10. (5 AND 9) 11. randomized controlled trial.pt. 12. controlled clinical trial.pt. 13. randomized.ab. 14. "randomized controlled trial".tw. 15. placebo.ab. 16. drug therapy.fs. 17. randomly.ab. 18. trial.ab. 19. groups.ab. 20. (11 OR 12 OR 13 OR 14 OR 15 OR 16 OR 17 OR 18 OR 19) 21. humans.sh. 22. (20 AND 21) 23. (10 AND 22) |
Searching other resources
We handsearched the reference lists of all relevant studies in order to identify any articles not found through the electronic search. We contacted the corresponding author of relevant studies to identify references and provide unpublished data. We approached pharmaceutical companies for additional published and unpublished trials.
Data collection and analysis
Selection of studies
Three review authors independently screened the titles and abstracts of citations obtained from the searches and discarded irrelevant articles. The three authors assessed retrieved studies for inclusion based on the pre‐determined criteria. We resolved any disagreements by consulting the co‐review authors.
Data extraction and management
Two authors independently extracted data from the included studies. We extracted numerical results of primary and secondary outcomes using a data extraction form developed by the review authors. We resolved any discrepancies through discussion or by consulting a third review author. One author entered information into Review Manager (RevMan) and the other authors checked the data for accuracy. In the case of multiple treatment groups (to test the effect of different dosages) in studies we included combined data forming a single group for the active treatment.
Assessment of risk of bias in included studies
Two authors assessed the methodological quality of each included study using the Cochrane Collaboration’s tool for assessing risk of bias (Higgins 2009). We assessed biases according to six criteria: sequence generation, allocation concealment, blinding, incomplete data outcomes, free from selective reporting and other bias. Each criterion was assigned a low risk of bias, a high risk of bias or an unclear risk of bias based on how adequately the study fulfilled each criterion.
Measures of treatment effect
We treated outcome measures that arose from ordinal scales (for example clinical global impression of change) as continuous variables. For continuous variables, we combined change scores from baseline and used them to determine mean difference (MD) with the 95% confidence interval (CI). Where change scores were unavailable, we extracted the mean, standard deviation and the number of people in each group at each time point. For dichotomous outcomes (that is adverse events (AE), attrition rates), we used risk ratio (RR) with 95% CI to estimate the treatment effect. In order to interpret findings, we assessed the overall quality of evidence for all outcomes with pooled data using the Grading of Recommendations Assessment, Development and Evaluation (GRADE) approach (Guyatt 2008). This gives an assessment of the confidence that can be placed in the effect estimate. Evidence from RCTs was initially considered 'high quality'. The rating was downgraded one level for each of: serious limitations in the study methodology (risk of bias), imprecision, indirectness in addressing our question of interest, inconsistencies in effect estimates (heterogeneity) and reporting or publication biases. Summary of findings tables were created using the GRADE profiler (GRADEpro) software.
Unit of analysis issues
For inclusion in this review, we analysed only the data from the first treatment phase for any studies employing a cross‐over design, and we analysed data from any subsequent phase separately. In the case of outcomes being reported at multiple time points, the primary endpoints (as stated by the study authors) were used in our primary pooled analysis. Other additional endpoints were included in subgroup analyses to examine the effect of treatment duration.
Dealing with missing data
We requested unreported data from the author of the original study and performed statistical calculations to complete the missing information, where possible. In the case of missing data for non‐completers, last observation carried forward (LOCF) data, performed by the study authors, were included in our analysis of treatment outcomes.
Assessment of heterogeneity
When pooling trials, we used the I2 statistic to identify heterogeneity across studies, where I2 > 40% indicated significant heterogeneity. If there was minimal heterogeneity, we used a fixed‐effect model analysis. In the case that heterogeneity was substantial, we conducted random‐effects model analyses. In addition, we performed analyses to determine sources of heterogeneity, including length of study, severity of AD and dose, if possible.
Assessment of reporting biases
We performed a comprehensive search, including trials registries, to minimize the risk of reporting biases.
Data synthesis
We combined study outcomes using a fixed‐effect model, using the inverse variance method to provide a pooled estimated effect from the data. In the case of significant heterogeneity, we used random‐effects models. All analyses were conducted according to the principles of intention to treat (ITT).
Subgroup analysis and investigation of heterogeneity
We conducted subgroup analyses based on the information available on duration of treatment. Doody 2008 evaluated treatment effects at 26 weeks with the option of blinded treatment for an additional 26 weeks. CONNECTION and CONCERT had a follow‐up period of 26 and 52 weeks, respectively. Therefore, subgroup analyses were completed in two parts: CONNECTION and Doody 2008 (26 weeks), and CONCERT and Doody 2008 (52 weeks). For the Doody 2008 study, LOCF was applied to participants who did not complete the total 52 weeks.
Sensitivity analysis
We did not conduct sensitivity analyses as there was a limited number of included studies.
Results
Description of studies
Results of the search
Our search of the database and other sources yielded 554 records after removing duplicates, 547 of which did not meet our inclusion criteria. In total, seven studies were assessed for eligibility. Five studies (CONCERT; CONNECTION; Doody 2008; NCT00838110; Tariot 2009) examined the effects of latrepirdine in the mild‐to‐moderate AD population while two studies (CONTACT; NCT00912288) investigated safety and efficacy in moderate‐to‐severe AD patients. One study had data published in a primary full‐text article (Doody 2008). One published the findings in a conference abstract (Tariot 2009). Two studies published their findings on safety and tolerability electronically on the ClinicalTrials.gov website (NCT00838110; NCT00912288). We specifically requested unreported data from the principal investigators and pharmaceutical companies (Pfizer and Medivation) for two of the included trials (CONCERT; CONNECTION) but did not receive a response. However, we were able to extract data reported in a meta‐analysis (Cano‐Cuenca 2014) that included results from CONCERT and CONNECTION. We also attempted to contact the authors of the meta‐analysis in order to obtain further information but did not receive a response. Both of the studies in patients with moderate‐to‐severe AD (CONTACT; NCT00912288) were terminated early by the sponsor. Some data on safety, but not efficacy, from NCT00912288 were available on the ClinicalTrials.gov website. No safety or efficacy results from CONTACT were available. We did not attempt to contact the investigators or sponsors of CONTACT, NCT00912288 and Tariot 2009 as these trials were also conducted by Medivation and Pfizer who failed to respond to our requests for information on CONNECTION and CONCERT. See Figure 1 for the study flow diagram.
Study flow diagram.
Included studies
We included seven studies in this review (see the Characteristics of included studies table).
We extracted information on efficacy measures from Doody 2008 and from the CONNECTION and CONCERT trials. No primary reports of data from CONNECTION and CONCERT have been published. We attempted to contact the studies' principal investigators but received no reply. Therefore, we extracted data for these two trials from a meta‐analysis by Cano‐Cuenca 2014, who was successful in obtaining the unpublished results. For analyses of safety and tolerability, we extracted data from the Doody 2008 paper, the Tariot 2009 conference presentation and results available on the clinicialtrials.gov website for the NCT00838110 and NCT00912288 trials. No information was available for the CONTACT study. Subgroup analyses were conducted to examine the effect of treatment duration on efficacy, safety and tolerability measures. We were not able to perform subgroup analyses for dosage as the studies we included combined the different treatment arms to form a single active treatment group.
Doody 2008 was the first randomized, double‐blind, placebo controlled phase II trial to examine the safety, tolerability and efficacy of latrepirdine in mild‐to‐moderate AD patients. All participants had a diagnosis of possible or probable AD according to the DSM‐IV (American Psychiatric Association 2013) and NINCDS‐ADRDA criteria (McKhann 1984). Patients taking ChEIs or memantine within 60 days of the time of randomization were excluded. Patients taking low doses of psychoactive drugs were permitted to enrol but were restricted per protocol. Enrolled patients had a caregiver who cared for them at least five days a week, agreed to attend study visits and provided information about the patient. All participants (MMSE (Folstein 1975) scores 10 to 24, age ≥ 50 years) were randomized to 60 mg (20 mg orally three times a day) of latrepirdine or matched placebo for 26 weeks. An additional six months of blinded treatment was also offered to participants who wished to continue in the study. LOCF analyses were carried out for those at 26 weeks who did not complete 52 weeks of the study treatment. In total, 89 patients received latrepirdine and 94 received placebo in the primary 26‐week treatment. Additionally, 120 patients (61 in the latrepirdine and 59 in the placebo group) continued and completed the extended 52‐week treatment. We used results from the 26‐week follow‐up in our primary pooled analysis as this was the stated primary endpoint for this study. The 52‐week results were used in subsequent subanalyses. The primary outcome measure was cognition (ADAS‐Cog (Weyer 1994)) at 26 weeks, while behaviour (NPI (Cummings 1994)), function (ADCS‐ADL (Galasko 1997)) and global assessment of change (CIBIC‐Plus (Reisberg 1997), ADCS‐CGIC (Schneider 1997)) were assessed as secondary outcome measures.
CONNECTION was a three‐arm randomized, double‐blind, placebo controlled phase III trial that was developed based on results from Doody 2008 to further investigate the safety and efficacy of latrepirdine in mild‐to‐moderate AD. Inclusion and exclusion criteria were similar between the two studies, except that patients taking ChEIs or memantine within 90 days of randomization were excluded from the CONNECTION trial. Participants were randomized to either 15 mg latrepirdine (5 mg orally three times a day), 60 mg latrepirdine (20 mg orally three times a day) or placebo for 26 weeks. In total, 200 patients were randomized to latrepirdine (combined active treatment arms) and 198 to placebo. Cano‐Cuenca 2014 did not indicate specific sample sizes within the 15 mg and 60 mg treatment groups. Primary efficacy outcome measures were cognition (ADAS‐Cog (Weyer 1994)) and global function (CIBIC‐Plus (Reisberg 1997)), while secondary outcome measures were behaviour (NPI (Cummings 1994)) and self care and function (ADCS‐ADL (Galasko 1997)).
CONCERT was also a three‐arm randomized, double‐blind, placebo controlled phase III trial derived from the initial Doody 2008 study to investigate the safety and efficacy of latrepirdine in mild‐to‐moderate AD over 12 months. However, enrolled patients were not required to have a caregiver who was largely involved in their care. All participants had a diagnosis of possible or probable AD according to the DSM‐IV‐TR (American Psychiatric Association 2013) and NINCDS‐ADRDA (McKhann 1984) criteria (MMSE scores 12 to 24) and were stable on donepezil for at least 6 months. Participants were randomized to either 15 mg latrepirdine (5 mg orally three times a day), 60 mg latrepirdine (20 mg orally three times a day) or placebo for 52 weeks. In total, 322 patients were randomized to latrepirdine (combined active treatment arms) and 340 patients to placebo. Cano‐Cuenca 2014 did not indicate specific sample sizes within the 15 mg and 60 mg treatment groups. Primary efficacy outcome measures were function (ADCS‐ADL (Galasko 1997)) and cognition (ADAS‐Cog (Weyer 1994)), and secondary outcome measures were global function (CIBIC‐Plus (Reisberg 1997)), behaviour (NPI (Cummings 1994)), resource utilization (Resource Utilization in Dementia questionnaire (RUD‐lite) (Wimo 2010)), and quality of life (EQ‐5D (Ankri 2003)).
NCT00912288 was a randomized, double‐blind, placebo controlled phase III trial that aimed to investigate the safety and efficacy of latrepirdine in AD patients in the moderate‐to‐severe stages of the disease. Eligible patients had a score of 5 to 14 on the MMSE, were stable on memantine for at least 6 months prior to randomization, and had a caregiver who assisted them at least 5 days per week for at least 3 hours per day. Participants were randomized to 60 mg (20 mg orally three times a day) of latrepirdine or matched placebo for 26 weeks. Primary efficacy outcome measures were cognition (SIB) and daily function (ADCS‐ADL), and secondary outcome measures were behaviour (NPI), global function (CIBIC‐Plus), resource utilization (RUD‐lite), quality of life (EQ‐5D) and adverse event reports. However, this study was terminated by the investigators on 7 May 2010 as a result of the lack of efficacy in the completed CONNECTION study. The investigators maintained that safety concerns were not the reason for early termination. In total, 86 patients were randomized (44 to latrepirdine and 42 to placebo). Five patients (3 on latrepirdine and 2 on placebo) completed the study before it was terminated. Efficacy data have not been made available but safety data were reported for follow‐ups of up to 30 weeks.
The phase III CONTACT was a similar study to NCT00912288. However, in this study patients needed to be on a stable dose of donepezil and not memantine. The primary outcome measures were behaviour (NPI) and daily function (ADCS‐ADL severe). In total, 89 patients were enrolled in the trial before it was terminated on 7 May 2010 due to the reason described above for the NCT00912288 trial.
NCT00838110 was a randomized, double‐blind, placebo controlled phase III trial associated with CONNECTION and CONCERT that aimed to evaluate the safety and tolerability of latrepirdine over 12 and 26 weeks. To meet the inclusion criteria participants needed to either not have taken anti‐dementia medication for at least 60 days or be on a stable dose of medications for at least 60 days prior to enrolment. Patients were excluded if they resided in a nursing home or assisted care facility and required 24‐hour care and supervision. The treatment arms included 60 mg latrepirdine (20 mg orally three times a day) and matched placebo. In total, 370 patients received latrepirdine and 371 received placebo for 26 weeks. Outcome measures included an assessment of vital signs, electrocardiogram, clinical chemistry and haemotology findings and adverse event reports.
Tariot 2009 was a randomized, double‐blind, placebo controlled phase I trial that aimed to explore the safety, tolerability and pharmacokinetics of latrepirdine in AD patients over 28 days. The investigators also compared different titration schedules. All participants had a diagnosis of possible or probable AD according to the DSM‐V (American Psychiatric Association 2013) and NINCDS‐ADRDA (McKhann 1984) criteria, although no MMSE criterion was imposed. Patients were required to be stable on 10 mg donepezil (and no other ChEI or memantine) for a minimum of 60 days. Patients on H1 receptor antihistamines, dextromethorphan or narcotic analgesics were excluded. Enrolled patients had a caregiver who cared for them at least five days a week. The treatment arms included 60 mg latrepirdine (20 mg orally three times a day) and matched placebo. The first 14 patients were placed on a gradual 3‐step titration (2.5 to 5 to 10 to 20 mg TID) and the final 10 patients were on a 2‐step rapid titration (10 to 20 mg three times a day). In total, 15 patients received latrepirdine and 9 received placebo for 28 days. Outcome measures included an assessment of vital signs, electrocardiogram, laboratory findings and adverse event reports.
Excluded studies
We did not exclude any RCTs of latrepirdine for mild, moderate or severe AD.
Risk of bias in included studies
All included trials were double‐blind, randomized, placebo controlled studies. Thus, we determined that all seven studies had adequate random sequence generation and blinding. Though Doody 2008 provided an adequate description of the treatment allocation concealment, we could not obtain this information for the other trials (CONCERT; CONNECTION; NCT00838110; NCT00912288; Tariot 2009). The missing outcome (attrition bias) reporting in Doody 2008, NCT00838110, NCT00912288 and Tariot 2009 were adequate. Those studies provided a description of attrition in their participant flow charts and dropout rates were comparable between the active treatment and placebo groups. We were not able to obtain information on participant attrition for CONTACT, CONNECTION and CONCERT and judged the available data from these trials to be at high risk of bias in this domain. Doody 2008 reported that the objectives specified in their protocol were met. All safety and tolerability specified outcome measures for NCT00838110 are reported in ClinicalTrials.gov. The Tariot 2009 conference presentation did not report values for all pre‐specified safety measures. The Cano‐Cuenca 2014 meta‐analysis did not report on some of the pre‐determined neuropsychological as well as safety and tolerability outcome measures stated in the CONNECTION and CONCERT protocols. CONTACT did not report data on any outcome measures. Thus, biases due to selective outcome reporting likely exists in these trials. See Characteristics of included studies table; Figure 2; and Figure 3.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Effects of interventions
See: Summary of findings for the main comparison Summary of findings for efficacy measures; Summary of findings 2 Summary of findings for safety and tolerability
Primary outcomes
Clinical global impression of change
Clinical global impression of change was assessed in Doody 2008, NCT00912288, CONCERT and CONNECTION using the CIBIC‐Plus. However, we were only able to extract results from Doody 2008. Higher scores on this assessment represent worsening (Reisberg 1997). The authors reported small but significant improvements on the CIBIC‐Plus for 183 patients (89 on latrepirdine and 94 on placebo) favouring latrepirdine following the 26‐week primary endpoint (MD ‐0.60, 95% CI ‐0.89 to ‐0.31, P < 0.001). Similar results were found at the additional 52‐week follow‐up (MD ‐0.70, 95% CI ‐1.01 to ‐0.39, P < 0.001). However, we considered this to be low quality evidence due to imprecision and publication bias. Thus, we could not draw conclusions about the efficacy of latrepirdine in terms of changes in clinical impression. See Analysis 1.1; Analysis 1.2; Figure 4; and Figure 5.
Forest plot of comparison: 1 Clinical global impression of change, outcome: 1.1 CIBIC‐Plus (change from baseline at 26 weeks).
Forest plot of comparison: 1 Clinical global impression of change, outcome: 1.2 CIBIC‐Plus (change from baseline at 52 weeks).
Cognition
CONCERT, CONNECTION and Doody 2008 all assessed cognition using the ADAS‐Cog and MMSE in patients with mild‐to‐moderate dementia. Higher scores on the ADAS‐Cog (Weyer 1994) represent worse performance while higher scores on the MMSE (Folstein 1975) represent better performance. The minimum clinically significant difference on the ADAS‐cog in mild AD has been considered to be approximately 3 to 4 points (Schrag 2012). We included 1243 patients (611 on latrepirdine and 632 on placebo) in our random‐effects model meta‐analysis. The result of the ADAS‐Cog analysis was compatible with either benefit or no effect of latrepirdine (MD ‐1.49, 95% CI ‐3.47 to 0.49, P = 0.14, 3 studies). There was considerable between‐study heterogeneity (I2 = 88%). We rated this as low quality evidence due to the imprecision and high inconsistency between studies. We also conducted subgroup analyses of results after 26 and 52 weeks. There was very significant heterogeneity in both these subgroup analyses (I2 = 94% and 93% respectively). At 26 weeks the MD was ‐1.97 (95% CI ‐5.78 to 1.84, n = 581, P = 0.31, 2 studies) and at 52 weeks the MD was ‐3.07 (95% CI ‐7.70 to 1.57, n = 845, P = 0.19, 2 studies). It was not possible to draw conclusions about the efficacy of latrepirdine from these subgroup analyses because of imprecision in the results. We considered this to be very low quality evidence because of the imprecision, serious inconsistency between studies and risk of bias. See Analysis 2.1; Analysis 2.2; Analysis 2.3; Figure 6; Figure 7; Figure 8.
Forest plot of comparison: 2 Cognition, outcome: 2.1 ADAS‐Cog (change from baseline).
Forest plot of comparison: 2 Cognition, outcome: 2.2 ADAS‐Cog (change from baseline at 26 weeks).
Forest plot of comparison: 2 Cognition, outcome: 2.3 ADAS‐Cog (change from baseline at 52 weeks).
Past studies have considered a 2 to 4 point change on the MMSE to be clinically relevant (Hensel 2007). Only Doody 2008 found significant improvements on MMSE scores. However, the combined data for 1243 patients were compatible with either benefit or no important effect of latrepirdine (MD 0.59, 95% CI ‐0.94 to 2.11, P = 0.45). Thus, the quality of evidence was downgraded to low as a result of imprecision and high heterogeneity (I2 = 92%) between studies. We found similar results for subgroup analyses at 26 weeks (MD 0.87, 95% CI ‐1.90 to 3.64, P = 0.54, 2 studies) and 52 weeks (MD 1.04, 95% CI ‐0.95 to 3.03, P = 0.31, 2 studies) with evidence of significant heterogeneity (I2 = 96% and 90% respectively). As a result of the serious inconsistency between studies as well as the imprecision and risk of bias, we rated the quality of evidence for these subgroup analyses to be low. See Analysis 2.4; Analysis 2.5; Analysis 2.6; Figure 9; Figure 10; Figure 11.
Forest plot of comparison: 2 Cognition, outcome: 2.4 MMSE (change from baseline).
Forest plot of comparison: 2 Cognition, outcome: 2.5 MMSE (change from baseline at 26 weeks).
Forest plot of comparison: 2 Cognition, outcome: 2.6 MMSE (change from baseline at 52 weeks).
Secondary outcomes
Function
Function was assessed using the ADCS‐ADL in CONCERT, CONNECTION and Doody 2008. Higher scores on this assessment indicate better performance (Galasko 1997). We found no clear evidence of benefit from latrepirdine in the pooled results (MD 1.00, 95% CI ‐1.15 to 3.15, P = 0.36, I2 = 74%, 3 studies), at 26 weeks (MD 1.81, 95% CI ‐1.01 to 4.63, P = 0.21, I2 = 77%) and 52 weeks (MD 1.84, 95% CI ‐3.00 to 6.68, P = 0.46, I2 = 89%). As evidence of considerable heterogeneity was found in all analyses for function, we used random‐effects models in our analyses. We rated the quality of evidence for the pooled analysis to be low due to imprecision and serious inconsistency across studies. As a result of serious inconsistency, imprecision and risk of bias we rated the quality of evidence for the subgroup analyses to be very low. See Analysis 3.1; Analysis 3.2; Analysis 3.3; Figure 12; Figure 13; Figure 14.
Forest plot of comparison: 3 Function, outcome: 3.1 ADCS‐ADL (change from baseline).
Forest plot of comparison: 3 Function, outcome: 3.2 ADCS‐ADL (change from baseline at 26 weeks).
Forest plot of comparison: 3 Function, outcome: 3.3 ADCS‐ADL (change from baseline at 52 weeks).
Behaviour
Overall behaviour was assessed using the NPI in CONCERT, CONNECTION and Doody 2008, with higher scores indicating increased behavioural disturbances (Cummings 1994). A 4 point change on the NPI has been considered clinically significant (Cummings 2000; Finkel 2004). The combined data for all three studies (n = 1243) revealed slight benefit with latrepirdine compared to placebo (MD ‐1.77, 95% CI ‐3.09 to ‐0.45, P = 0.009) although possibly not enough to be of clinical significance. There was low heterogeneity in this analysis (I2 = 32%). Thus we considered this to be high quality evidence. Subgroup analysis at 26 weeks for 581 patients also suggested greater improvement with latrepirdine than placebo (MD ‐2.51, 95% CI ‐4.24 to ‐0.79, P = 0.004, 2 studies) with low heterogeneity (I2 = 17%). However, we considered this to be low quality evidence due to imprecision as well as risk of bias. We found a trend towards benefit of latrepirdine on NPI compared with placebo at 52 weeks (MD ‐1.61, 95% CI ‐3.27 to 0.06, P = 0.06, 2 studies) with moderate heterogeneity (I2 = 53%). As inconsistency between studies was not serious we rated this to be moderate quality evidence due only to risk of bias. See Analysis 4.1; Analysis 4.2; Analysis 4.3; Figure 15; Figure 16; Figure 17.
Forest plot of comparison: 4 Behaviour, outcome: 4.1 NPI (change from baseline).
Forest plot of comparison: 4 Behaviour, outcome: 4.2 NPI (change from baseline at 26 weeks).
Forest plot of comparison: 4 Behaviour, outcome: 4.3 NPI (change from baseline at 52 weeks).
Adverse events
Doody 2008 found that dry mouth and depressed mood or depression were more prevalent in patients on latrepirdine when compared with placebo. However, the combined data for Doody 2008, NCT00912288, NCT00838110 and Tariot 2009, with 1034 patients (518 on latrepirdine and 516 on placebo), indicated that there was no significant difference between active treatment and placebo in the number of adverse events experienced by patients (risk ratio (RR) 1.03, 95% CI 0.93 to 1.14, P = 0.51) with evidence of low heterogeneity (I2 = 35%). Similarly, serious adverse events, described as either severe or life‐threatening, were not significantly different between the latrepirdine and placebo groups (RR 0.86, 95% CI 0.55 to 1.35, P = 0.52) with no evidence of heterogeneity. We rated the quality of evidence to be moderate due to possible publication bias. See Analysis 5.1; Analysis 5.2; Figure 18; Figure 19.
Forest plot of comparison: 5 Adverse events, outcome: 5.1 Adverse events.
Forest plot of comparison: 5 Adverse events, outcome: 5.2 Serious adverse events.
Dropouts
Combined data for Doody 2008, NCT00912288, NCT00838110 and Tariot 2009 indicated comparable attrition rates between treatment groups (RR 0.91, 95% CI 0.65 to 1.27, P = 0.57) for 1034 AD patients, with no heterogeneity. Furthermore, there were no significant differences in dropouts due to adverse events (RR 0.98, 95% CI 0.57 to 1.67, P = 0.93, 3 studies) for 1010 patients, with no heterogeneity. We considered the evidence for attrition in these studies to be moderate due to publicaion bias. See Analysis 6.1; Analysis 6.2; Figure 20; Figure 21.
Forest plot of comparison: 6 Dropouts, outcome: 6.1 Total dropouts.
Forest plot of comparison: 6 Dropouts, outcome: 6.2 Dropouts due to adverse events.
Discussion
Summary of main results
Overall, this meta‐analysis of seven studies suggests that latrepirdine does not provide significant clinical, cognitive or functional benefits to patients with mild‐to‐moderate AD. However, we found modest improvements in behavioural symptoms in those treated with latrepirdine at 26 weeks and a trend towards significant improvement in those treated with latrepirdine at 52 weeks. Latrepirdine was both safe (frequency of adverse events and serious adverse events) and well tolerated (overall attrition rates and attrition due to adverse events) compared with placebo.
Overall completeness and applicability of evidence
By conducting a meta‐analysis, we were able to combine the findings from multiple clinical trials to evaluate the efficacy of latrepirdine in the treatment of patients with AD. When evaluating the applicability of this review to the clinical setting, outcome measures, dementia stage, treatment duration and adherence should be taken into consideration.
Overall, we were able to analyse most of our chosen outcome measures, which are the standard assessments currently used in clinical practice. However, we were not able to obtain results on important clinical outcomes from the CONTACT, CONCERT and CONNECTION trials, including the ADCS‐CGIC and CIBIC‐Plus. As well, the included studies did not report information on many of our pre‐determined secondary outcomes, such as mood, quality of life, caregiver burden and institutionalization. All studies included in this review were double‐blind, placebo controlled RCTs targeting patients with mild‐to‐moderate and moderate‐to‐severe AD. The two trials (CONTACT; NCT00912288) of moderate‐to‐severe patients did not report primary outcome measures. Thus, there are no data available to evaluate the efficacy of latrepirdine in the more severe stages of dementia. Study duration ranged from 12 weeks to 52 weeks, with a mean duration of 26 weeks. This variation in treatment duration allowed us to assess the longitudinal effects of latrepirdine, which are of value when applied to clinical practice. In addition, Doody 2008 and CONNECTION randomized patients with MMSE scores between 10 and 24, while CONCERT recruited patients with MMSE scores of 12 to 24 and the NCT00838110 trial recruited patients with MMSE scores of 12 to 26. NCT00912288 examined patients with MMSE scores of 5 to 14. Tariot 2009 did not implement a strict MMSE cut‐off. These discrepancies in MMSE inclusion criteria between studies may have resulted in heterogeneity. Finally, treatment adherence in RCTs is often more standardized and monitored more thoroughly than in real‐world clinical settings. However, measures of adherence to the drug were not described and results on adherence rates were not reported in any of the included studies. This may limit our ability to assess the external validity of the results in terms of clinical settings.
Quality of the evidence
The quality of evidence for our primary outcome measures was low to moderate, while the quality of evidence for our subgroup analyses was very low to low. Overall, quality was downgraded due to inconsistency between studies, imprecision and publication bias. Although six studies were included in the quantitative meta‐analysis, only published data were extracted for analysis since efforts to contact the investigators of CONCERT and CONNECTION as well as the Cano‐Cuenca 2014 study did not yield a response. Additionally, no results from the CONTACT trial have been published. Further, the strong evidence of benefit found in Doody 2008 may be biasing the overall synthesized effect in our meta‐analysis, particularly since we had a small number of studies available for analysis. As a result, while this review was able to address all primary outcome measures it was done in a limited way. For the unattainable outcome measures of this review, we will modify results pending the publication of data from future studies.
As heterogeneity was high in most outcome measures we should take into consideration the potential sources of heterogeneity, which may have an impact on the quality of the evidence. The CONCERT trial required patients to be on a stable dosage of donepezil for at least six months, Tariot 2009 required patients to be stable on 10 mg donepezil for 60 days, and NCT00838110 allowed patients to either be free of or on stable anti‐dementia medications for at least 60 days. Patients in NCT00912288 were taking memantine for at least six months prior to enrolment. Doody 2008 and CONNECTION excluded patients if they had taken anti‐dementia therapy within the 60 days prior to dosing. Doody 2008, CONNECTION and NCT00912288 also required that patients had a caregiver who was involved in the patient’s care for at least five days a week, while this was not the case for CONCERT or the NCT00838110 trial. This may have had an effect on scores of the outcome measures which require caregiver information. In addition, the NCT00838110 trial excluded any institutionalized patients with need for 24‐hour care and supervision. Finally, CONCERT and CONNECTION had two treatment arms consisting of a 15 mg and 60 mg dose of treatment while Doody 2008 only used the 60 mg dose. We were not able to explore the effect of dosing schedule since the Cano‐Cuenca 2014 meta‐analysis did not detail the sample sizes within the active treatment arms for these two trials. Thus, discrepancies in use of concomitant anti‐dementia medications, study duration, caregiver requirements, institutionalization and dosing are all potential sources of heterogeneity which may limit the quality of the evidence.
Potential biases in the review process
By performing a meta‐analysis and a systematic review, we were able to evaluate the pooled data from existing clinical trials of latrepirdine in patients with AD. It is possible that there have been other clinical trials conducted that have not been published as a result of negative findings. For instance, we could not evaluate latrepirdine in moderate‐to‐severe patients as the one trial (CONTACT) aiming to explore this question was terminated with no data provided. Ideally, we planned to include data from the completed CONNECTION and CONCERT trials in the evaluation of our primary and secondary outcome measures. However, since data were not made available from their respective investigators upon request, the findings of this meta‐analysis reflect published data only.
Agreements and disagreements with other studies or reviews
The findings of our review are in agreement with the meta‐analysis conducted by Cano‐Cuenca 2014; latrepirdine should not be recommended as an anti‐dementia treatment. The review by Cano‐Cuenca 2014 included three of the clinical trials in our review; Doody 2008, CONNECTION and CONCERT. Cano‐Cuenca 2014 determined that although latrepirdine was well tolerated, both alone and when given in combination with other drugs, there was only a modest but significant effect found on behaviour and no effect on cognition in AD patients. This is in apparent agreement with the pharmaceutical company who was developing this medication as the results of the CONNECTION and CONCERT studies have not been published, they declined to provide any results, and to our knowledge they have not pursued this indication.
Risk of bias graph: review authors' judgements about each risk of bias item presented as percentages across all included studies.
Risk of bias summary: review authors' judgements about each risk of bias item for each included study.
Forest plot of comparison: 1 Clinical global impression of change, outcome: 1.1 CIBIC‐Plus (change from baseline at 26 weeks).
Forest plot of comparison: 1 Clinical global impression of change, outcome: 1.2 CIBIC‐Plus (change from baseline at 52 weeks).
Forest plot of comparison: 2 Cognition, outcome: 2.1 ADAS‐Cog (change from baseline).
Forest plot of comparison: 2 Cognition, outcome: 2.2 ADAS‐Cog (change from baseline at 26 weeks).
Forest plot of comparison: 2 Cognition, outcome: 2.3 ADAS‐Cog (change from baseline at 52 weeks).
Forest plot of comparison: 2 Cognition, outcome: 2.4 MMSE (change from baseline).
Forest plot of comparison: 2 Cognition, outcome: 2.5 MMSE (change from baseline at 26 weeks).
Forest plot of comparison: 2 Cognition, outcome: 2.6 MMSE (change from baseline at 52 weeks).
Forest plot of comparison: 3 Function, outcome: 3.1 ADCS‐ADL (change from baseline).
Forest plot of comparison: 3 Function, outcome: 3.2 ADCS‐ADL (change from baseline at 26 weeks).
Forest plot of comparison: 3 Function, outcome: 3.3 ADCS‐ADL (change from baseline at 52 weeks).
Forest plot of comparison: 4 Behaviour, outcome: 4.1 NPI (change from baseline).
Forest plot of comparison: 4 Behaviour, outcome: 4.2 NPI (change from baseline at 26 weeks).
Forest plot of comparison: 4 Behaviour, outcome: 4.3 NPI (change from baseline at 52 weeks).
Forest plot of comparison: 5 Adverse events, outcome: 5.1 Adverse events.
Forest plot of comparison: 5 Adverse events, outcome: 5.2 Serious adverse events.
Forest plot of comparison: 6 Dropouts, outcome: 6.1 Total dropouts.
Forest plot of comparison: 6 Dropouts, outcome: 6.2 Dropouts due to adverse events.
Comparison 1 Clinical global impression of change, Outcome 1 CIBIC‐Plus (change from baseline at 26 weeks).
Comparison 1 Clinical global impression of change, Outcome 2 CIBIC‐Plus (change from baseline at 52 weeks).
Comparison 2 Cognition, Outcome 1 ADAS‐Cog (change from baseline).
Comparison 2 Cognition, Outcome 2 ADAS‐Cog (change from baseline at 26 weeks).
Comparison 2 Cognition, Outcome 3 ADAS‐Cog (change from baseline at 52 weeks).
Comparison 2 Cognition, Outcome 4 MMSE (change from baseline).
Comparison 2 Cognition, Outcome 5 MMSE (change from baseline at 26 weeks).
Comparison 2 Cognition, Outcome 6 MMSE (change from baseline at 52 weeks).
Comparison 3 Function, Outcome 1 ADCS‐ADL (change from baseline).
Comparison 3 Function, Outcome 2 ADCS‐ADL (change from baseline at 26 weeks).
Comparison 3 Function, Outcome 3 ADCS‐ADL (change from baseline at 52 weeks).
Comparison 4 Behaviour, Outcome 1 NPI (change from baseline).
Comparison 4 Behaviour, Outcome 2 NPI (change from baseline at 26 weeks).
Comparison 4 Behaviour, Outcome 3 NPI (change from baseline at 52 weeks).
Comparison 5 Adverse events, Outcome 1 Adverse Events.
Comparison 5 Adverse events, Outcome 2 Serious Adverse Events.
Comparison 6 Dropouts, Outcome 1 Total Drop‐out.
Comparison 6 Dropouts, Outcome 2 Drop‐out due to Adverse Events.
| Latrepirdine compared with placebo in Alzheimer's disease | ||||||
| Patient or population: patients with Alzheimer's disease | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Control | Latrepirdine | |||||
| CIBIC‐Plus (change from baseline at 26 weeks) | The mean CIBIC‐Plus (change from baseline at 26 weeks) in the control groups was 4.30 | The mean CIBIC‐Plus (change from baseline at 26 weeks) in the intervention groups was | 183 | ⊕⊕⊝⊝ | ||
| ADAS‐Cog | The mean change from baseline in ADAS‐Cog ranged across control groups from | The mean change from baseline in ADAS‐Cog in the intervention groups was 1.49 lower (95% CI ‐3.47 to 0.49) | 1243 | ⊕⊕⊝⊝ | 3 to 4 point change suggested to be clinically significant | |
| MMSE | The mean change from baseline in MMSE ranged across control groups from ‐1.90 to 1.19 | The mean change from baseline in MMSE in the intervention groups was 0.59 higher (95% CI ‐0.94 to 2.11) | 1243 | ⊕⊕⊝⊝ | 2 to 4 point change suggested to be clinically significant | |
| ADCS‐ADL | The mean change from baseline in ADCS‐ADL ranged across control groups from ‐6.91 to ‐0.74 | The mean change from baseline in ADCS‐ADL in the intervention groups was 1.00 higher (95% CI ‐1.15 to 3.15) | 1243 | ⊕⊕⊝⊝ | Limited data on clinically meaningful changes | |
| NPI | The mean change from baseline in NPI ranged across control groups from ‐1.04 to 3.65 | The mean change from baseline in NPI in the intervention groups was 1.77 lower (95% CI 3.09 to ‐0.45) | 1243 | ⊕⊕⊕⊕ | 4‐point change suggested to be clinically significant | |
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 significant imprecision 3 significant heterogeneity across studies | ||||||
| Latrepirdine compared with placebo in measures of safety and tolerability in Alzheimer's disease | ||||||
| Patient or population: patients with Alzheimer's disease | ||||||
| Outcomes | Illustrative comparative risks* (95% CI) | Relative effect | No of Participants | Quality of the evidence | Comments | |
| Assumed risk | Corresponding risk | |||||
| Control | Latrepirdine | |||||
| Adverse events | Study population | RR 1.03 | 1034 | ⊕⊕⊕⊝ | ||
| 583 per 1000 | 601 per 1000 | |||||
| Moderate | ||||||
| 569 per 1000 | 586 per 1000 | |||||
| Serious adverse events | Study population | RR 0.86 | 1034 | ⊕⊕⊕⊝ | ||
| 74 per 1000 | 63 per 1000 | |||||
| Moderate | ||||||
| 73 per 1000 | 63 per 1000 | |||||
| Total dropouts | Study population | RR 0.91 | 1034 | ⊕⊕⊕⊝ | ||
| 122 per 1000 | 111 per 1000 | |||||
| Moderate | ||||||
| 115 per 1000 | 105 per 1000 | |||||
| Dropouts due to adverse events | Study population | RR 0.98 | 1010 | ⊕⊕⊕⊝ | ||
| 51 per 1000 | 50 per 1000 | |||||
| Moderate | ||||||
| 46 per 1000 | 45 per 1000 | |||||
| *The basis for the assumed risk (e.g. the median control group risk across studies) is provided in footnotes. The corresponding risk (and its 95% confidence interval) is based on the assumed risk in the comparison group and the relative effect of the intervention (and its 95% CI). | ||||||
| GRADE Working Group grades of evidence | ||||||
| 1 publication bias | ||||||
| Source | Search Strategy |
| Medline (Ovid SP) | 1. latriperdine.mp. 2. dimebon.mp. 3. dimebolin.mp. 4. 2,3,4,5‐tetrahydro‐2,8‐dimethyl‐5‐(2‐(6‐methyl‐3‐pyridyl)ethyl)‐1H‐pyrido(4,3‐b)indole.mp. 5. (1 OR 2 OR 3 OR 4) 6. (Alzheimer* OR AD).mp. 7. Alzheimer Disease/ 8. Dement*.mp. 9. (6 OR 7 OR 8) 10. (5 AND 9) 11. randomized controlled trial.pt. 12. controlled clinical trial.pt. 13. randomized.ab. 14. "randomized controlled trial".tw. 15. placebo.ab. 16. drug therapy.fs. 17. randomly.ab. 18. trial.ab. 19. groups.ab. 20. (11 OR 12 OR 13 OR 14 OR 15 OR 16 OR 17 OR 18 OR 19) 21. humans.sh. 22. (20 AND 21) 23. (10 AND 22) |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 CIBIC‐Plus (change from baseline at 26 weeks) Show forest plot | 1 | 183 | Mean Difference (IV, Fixed, 95% CI) | ‐0.60 [‐0.89, ‐0.31] |
| 2 CIBIC‐Plus (change from baseline at 52 weeks) Show forest plot | 1 | 183 | Mean Difference (IV, Fixed, 95% CI) | ‐0.70 [‐1.01, ‐0.39] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 ADAS‐Cog (change from baseline) Show forest plot | 3 | 1243 | Mean Difference (IV, Random, 95% CI) | ‐1.49 [‐3.47, 0.49] |
| 2 ADAS‐Cog (change from baseline at 26 weeks) Show forest plot | 2 | 581 | Mean Difference (IV, Random, 95% CI) | ‐1.97 [‐5.78, 1.84] |
| 3 ADAS‐Cog (change from baseline at 52 weeks) Show forest plot | 2 | 845 | Mean Difference (IV, Random, 95% CI) | ‐3.07 [‐7.70, 1.57] |
| 4 MMSE (change from baseline) Show forest plot | 3 | 1243 | Mean Difference (IV, Random, 95% CI) | 0.59 [‐0.94, 2.11] |
| 5 MMSE (change from baseline at 26 weeks) Show forest plot | 2 | 581 | Mean Difference (IV, Random, 95% CI) | 0.87 [‐1.90, 3.64] |
| 6 MMSE (change from baseline at 52 weeks) Show forest plot | 2 | 845 | Mean Difference (IV, Random, 95% CI) | 1.04 [‐0.95, 3.03] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 ADCS‐ADL (change from baseline) Show forest plot | 3 | 1243 | Mean Difference (IV, Random, 95% CI) | 1.00 [‐1.15, 3.15] |
| 2 ADCS‐ADL (change from baseline at 26 weeks) Show forest plot | 2 | 581 | Mean Difference (IV, Random, 95% CI) | 1.81 [‐1.01, 4.63] |
| 3 ADCS‐ADL (change from baseline at 52 weeks) Show forest plot | 2 | 845 | Mean Difference (IV, Random, 95% CI) | 1.84 [‐1.00, 6.68] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 NPI (change from baseline) Show forest plot | 3 | 1243 | Mean Difference (IV, Fixed, 95% CI) | ‐1.77 [‐3.09, ‐0.45] |
| 2 NPI (change from baseline at 26 weeks) Show forest plot | 2 | 581 | Mean Difference (IV, Fixed, 95% CI) | ‐2.51 [‐4.24, ‐0.79] |
| 3 NPI (change from baseline at 52 weeks) Show forest plot | 2 | 845 | Mean Difference (IV, Fixed, 95% CI) | ‐1.61 [‐3.27, 0.06] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Adverse Events Show forest plot | 4 | 1034 | Risk Ratio (M‐H, Fixed, 95% CI) | 1.03 [0.93, 1.14] |
| 2 Serious Adverse Events Show forest plot | 4 | 1034 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.86 [0.55, 1.35] |
| Outcome or subgroup title | No. of studies | No. of participants | Statistical method | Effect size |
| 1 Total Drop‐out Show forest plot | 4 | 1034 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.91 [0.65, 1.27] |
| 2 Drop‐out due to Adverse Events Show forest plot | 3 | 1010 | Risk Ratio (M‐H, Fixed, 95% CI) | 0.98 [0.57, 1.67] |
