Aligning the evidence with practice: NICE guidelines for drug treatment of Alzheimer’s disease

Keywords::

• Alzheimer’s disease
• cholinesterase inhibitors
• cost–effectiveness
• National Institute for Clinical Excellence (UK)

There are an estimated 35.6 million people with dementia worldwide [1]. More than half of these individuals have Alzheimer’s disease (AD), resulting in a progressive decline in cognition, function and communication, together with the frequent occurrence of neuropsychiatric symptoms. In addition to the enormous human costs to the individuals and their families, dementia costs the global economy GB£604 billion a year [1]. There is a pressing clinical and economic need to find treatments that are both effective and cost effective.

Cholinesterase inhibitors, licensed in the late 1990s for mild–moderate AD, were the first widely available treatments. These drugs had a significant impact on treatment and care, giving hope to millions of people with AD and resulting in widespread changes in clinical services to facilitate earlier diagnosis and treatment. Whilst the ancillary benefits for the delivery of clinical care are important, the most crucial issue is whether these therapies actually provide an effective pharmacological treatment for people with AD.

There are more than 30 randomized controlled trials (RCTs) of the cholinesterase inhibitors donepezil (Aricept^®), rivastigmine (Exelon^®) and galantamine (Reminyl^®) in people with mild–moderate AD [2]. They confer significant benefits in cognition, function and global clinical outcome, with approximately half of the studies demonstrating additional benefits in neuropsychiatric symptoms [3]. These benefits are broadly similar for the different cholinesterase inhibitors. The usual pattern is for mean cognitive scores to improve and remain above pretreatment levels for up to 9 months [2]. The overall magnitude of benefit compared with placebo corresponds to a small–moderate Cohen’s D standardized effect size of approximately 0.4. Four lines of evidence have been put forward to suggest that this is an important and clinically meaningful benefit. First, there is clear benefit for cholinesterase inhibitors with respect to global clinical outcome [3–5]. Second, ‘buying’ patients up to 9 months of improved cognition has to be understood in the context of a disease where life expectancy from diagnosis is between 4 and 5 years [6]. Third, audit studies have demonstrated significant benefits in 40–70% of individuals treated with cholinesterase inhibitors [7]. Fourth, and most important, a pivotal RCT using Goal Attainment Scaling as the primary outcome measure demonstrated significant benefits for the cholinesterase inhibitor galantamine compared with placebo on outcome measures prioritized by the patients themselves [8].

The value of longer-term treatment is less clear, but open-label extension studies and one 2-year RCT provide some evidence for continuing benefit [9–11]. Although most studies have focused on mild–moderate AD, several RCTs with donepezil have suggested a similar magnitude of benefit for people with severe AD [12], and the license for donepezil has now been extended accordingly. One of the cholinesterase inhibitors, rivastigmine, is now available in a transdermal patch [13]. This delivery method appears to have a similar efficacy but significantly improved tolerability, and may have advantages for treatment compliance.

The first NICE guidance relating to anticholinesterase inhibitors was published in 2001. It recommended that all three drugs should be available to people with mild–moderate AD. A review in 2003 was extended to include memantine (Ebixa^®). The 2005 draft guidance recommended that none of the four drug treatments should be prescribed as NHS therapies because, while they were clinically effective, according to the NICE model they were not cost effective. However, following a massive wave of criticism, NICE sought further evidence that enabled the exploration of the efficacy of the treatments in particular subgroups of patients. Subgroups based on severity were examined and the drugs were found to have different cost–effectiveness profiles, although they were clinically effective in both mild and moderate stages. The final 2006 guidance stated that anticholinesterase inhibitors should be prescribed to individuals in the moderate stage of AD, as defined by a Mini-Mental State Examination (MMSE) score of between 10 and 20.

The 2006 recommendations were heavily criticized. Individuals with AD and their carers feel very strongly that the goal of anticholinesterase treatment should be to ‘buy time’ during the period when symptoms are mild. This is when people are best able to cope with symptoms and retain some independence and quality of life. Delaying treatment until the moderate stages also contradicts important policy aims of increasing rates of diagnosis and early intervention. Individuals experiencing symptoms often delay seeking help, and general practitioners (GPs) can be reluctant to diagnose dementia in the early stages. The absence of licensed drug treatments for the early stages is likely to exacerbate this issue.

The NICE guidance was further condemned for using the MMSE as a tool to assess suitability for drug treatment due to its poor inter-rater and test–retest reliability [14]. In addition, it is well recognized that race, education and language ability can affect performance on MMSE [15,16]. Criticisms were also made of the economic model. In particular, outcome measures prioritized by carers were not incorporated. The decision to withhold treatment from people with mild AD on the basis of possible differences in cost–effectiveness reflected a failure to model the pattern of cognitive decline across the disease, and an over-reliance on admission to care homes, a late-stage cost-driver. The length of time between mild AD and the need for full-time care requires considerable extrapolation to calculate the probable impact on cost. In addition, the calculation of cost–effectiveness did not capture important differences in the cost of care and quality of life between mild and moderate stages. This perspective is supported by the similar response rates in people with mild and moderate AD (responder vs nonresponder, based upon the NICE criteria).

A review of the 2006 NICE guidance is now underway incorporating a systematic review of cost and clinical effectiveness and a revised economic model. The new model contains some significant improvements. While the key outcome measure remains the delay in entry to full-time care, it does acknowledge the heterogeneity of costs and quality of life in the pre-full-time care state. However, the evaluation fails to incorporate the benefits the drugs provide to carers, which is partly due to a lack of data. Given the significant burden on carers of people with dementia and the increased risk of psychological morbidity and reduced quality of life [17], it is important that methods for incorporating benefit to carers are developed.

The updated systematic review and revised model have resulted in two major proposed changes. Treatment would be extended to people with mild and severe AD, as well as those in the moderate stage. This rationale is based on evidence of clinical effectiveness of early treatment and a common sense argument whereby treatment in the early stages will buy people more time when they are most independent. It will also encourage earlier diagnosis by GPs. An alternative economic model, submitted by the manufacturer of donepezil, used a discrete event simulation approach over a lifetime and demonstrated that the drugs are cost effective at both stages. The cost–effectiveness of treatment for people with more severe AD is less clear-cut, although it is helpful to allow clinicians more flexibility in the pharmacological treatment of these individuals, and there may be some secondary benefits related to reduced prescribing of psychotropic medications, including harmful antipsychotics.

The final guidance is expected in late 2010 or early 2011. Should it remain unchanged, it will provide a treatment option for individuals with AD upon diagnosis. The audit support tool for the 2006 guidance made clear that all people with moderate AD should be considered for treatment with one of the Alzheimer’s drugs. It is expected that a similar standard will be required in the forthcoming guidance, requiring considerable work to increase rates of diagnosis. The National Audit Office found that in England, only a third of people ever receive a formal diagnosis [18]. An international comparison found that the UK was in the bottom third of EU countries, even before access was restricted to people with an MMSE of 10–20 [101]. It is hoped that ongoing work to raise awareness and improve early diagnosis within the National Dementia Strategy for England will raise diagnosis rates.

An additional consideration in the UK is the recent announcement of plans to devolve power to local GP consortia to decide which drugs should be funded by the NHS [19]. These consortia will take the place of Primary Care Trusts by 2013, and NICE guidance will be advisory. This may increase the flexibility of clinicians to prescribe as they see fit, but there is a risk of access to drugs being determined by geography, as was the case for people with Alzheimer’s before the publication of the first NICE guidance.

Internationally, more rigorous evaluation of cost–effectiveness is becoming an increasing reality. The NICE approach to cost–effectiveness has been influential beyond England, and the current NICE evaluation of antidementia treatments has been closely scrutinized. Encouragingly, the economic model for evaluating treatments in people with dementia has improved, although further developments are still needed, ideally with a robust expert consensus to establish an optimal approach based on current knowledge. A clear demonstration of the cost–effectiveness of cholinesterase inhibitors through the current NICE appraisal will play an important role in developing better treatment and care in England, and is likely to have a positive influence on decision-making much more widely.

Financial & competing interests disclosure

Within the last 5 years, Clive Ballard has received honoraria from Novartis and Eisai Pharmaceutical Companies who have an interest in cholinesterase inhibitor therapies. The authors have no other relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript apart from those disclosed.

No writing assistance was utilized in the production of this manuscript.