This two-step procedure, which relies on the Diagnostic and Statistic Manual of Mental Disorders IV Text Revision (DSM-IV-TR) and the National Institute of Neurological and Communication Disorders and Stroke – Alzheimer’s Disease and Related Disorders Association (NINCDS-ADRDA) criteria, should be revised on the basis of several arguments. First, the criteria do not take into account the unprecedented growth of scientific knowledge concerning the existence of reliable biomarkers of AD that are now available through structural magnetic resonance imaging (MRI), molecular neuroimaging and cerebrospinal fluid analyses. Nor do they take into account the AD phenotype, which presents in most cases as a progressive amnestic dementia related to other Alzheimer’s-related changes that involve the medial temporal structures early in the course of disease. Furthermore, the episodic memory disorders of AD correlate well with a distribution of neurofibrillary tangles within the medial temporal lobe (MTL) and with the demonstration by MRI of volumetric loss of the hippocampus, the structure known to be critical for episodic memory.

In addition, recently developed disease-modifying therapies require early intervention at the prodromal stage before full-blown dementia. At the moment, the prodromal stage of AD is included under the heterogeneous term mild cognitive impairment (MCI). This heterogeneity may have contributed to the negative outcomes of clinical trials in which none of the drugs was successful in delaying the time to diagnosis of AD. It may be assumed that the heterogeneity of MCI has diluted the potential for a significant treatment effect, particularly considering that AD is already at work on the brain long before the onset of clinical dementia. However, it is possible to recognise this pre-dementia stage of AD by adopting a multidimensional approach, identifying:

• a specific amnestic disorder of the hippocampal type;
• the atrophy of medial temporal structures – specifically the hippocampus; and
• the specific profile of cerebrospinal fluid biomarkers or of metabolic neuroimaging changes.

An international working group was convened to discuss the opportunity to develop a diagnostic framework for AD that would include the prodromal stages. At the end of this consensus meeting it was concluded that it was possible to recognise AD at the prodromal, pre-dementia stage with the use of specific memory tests, biomarkers and neuroimaging investigations. There was no longer a reason to limit the diagnosis of AD to patients who reached the threshold of fullblown dementia. Accordingly, it was decided that new criteria be proposed that would apply both in the early stages and across the full spectrum of the illness.

Proposed Diagnostic Criteria for Probable Alzheimer’s Disease The framework addresses the presentations that are typical of AD. It excludes atypical presentations – primary progressive aphasia and visuospatial dysfunction – although it has been demonstrated that these atypical phenotypes can be associated with post mortem AD histological changes. To meet criteria for probable AD, an affected individual must fulfil the core clinical criterion (criterion A) and at least one of the supportive biomarker criteria (see Table 1).

To satisfy criterion A, memory symptoms must start gradually and show progressive decline over at least six months. Particular attention should be paid to intra-individual decline, which improves the identification of those individuals with prodromal AD. The proposed criteria emphasise the specificity of memory changes of AD and the need to use specific memory tests. It is noteworthy that most of the current memory tests do not record whether items to be recalled have been truly registered. Effective encoding of information should be controlled in order to exclude memory deficit related to anxiety, depression, frontal dysfunction or any other functional disorder. In the same way, identification of AD can be improved by using semantic cueing that facilitates the retrieval of stored information in aged healthy people or in patients with subcorticofrontal dysfunction. Reduced benefit of cueing at recall reliably identifies prodromal AD. Episodic memory impairment is proposed as a core feature of AD. It can be isolated or associated with other cognitive changes at the onset of AD or as AD advances. As AD advances, these changes become notable and can involve several domains – executive function, language, praxis, complex visual processing and gnosis. The emergence of neuropsychiatric symptoms, including apathy or delusions, also constitutes a clinical marker of the disease. However, even in these more advanced cases there should be evidence of an early and previous episodic memory deficit as a mandatory requirement for the diagnosis of AD.