A standardised, multilingual, Mini Linguistic State Examination (MLSE) for classifying and monitoring Primary Progressive Aphasia

Dementia is a complex and multifaceted clinical state, which shows marked variations in the biological mechanisms and molecular structure of brain pathology, and in the effects that these pathologies have on brain function. Importantly, not all people with dementia are suffering from Alzheimer's disease (though this remains the commonest underlying pathology), and not all experience forgetfulness as the first sign of mental difficulty.

One type of dementia presentation that has been extensively studied over recent years is one that begins with selective difficulty understanding and/or producing speech and language - a pattern known as primary progressive aphasia (PPA). Because language is such a complex mental ability, and relies on a number of widely distributed brain areas, PPA can take a variety of forms. Currently, three distinct patterns are recognised. Some patients experience a reduced ability to appreciate the meaning of words or objects ('semantic dementia'), while in others the problems lie in the production of words and sentences ('progressive nonfluent aphasia'). A third group displays milder problems, along with a striking inability to repeat spoken sentences ('logopenic progressive aphasia').

At the level of pathology these three dementia syndromes are far from distinct, making their biological basis difficult to determine during life. Improved understanding should lead to the development of disease specific treatments, but because of their relative rarity (less than 5% of cases of dementia begin with PPA), clinical trials will necessitate large scale, international collaborations, and these will inevitably include patients who speak languages other than English. Such a diversity of language communities will obviously present problems in the context of a condition that primarily affects language itself. The principal aim of this project, therefore, is to lay the foundations of a common descriptive currency, in which the three types of PPA can be described, and patients' abilities quantified on the same scale, regardless of the language in which they are assessed.

To achieve this, we will develop a brief language assessment instrument, which will allow the three PPA syndromes to be distinguished from one another on a common set of criteria and monitored over time using equivalent levels of severity. In view of the worldwide impact and importance of the Mini Mental State Examination (MMSE) in defining and quantifying dementia more generally, this brief language test will be given the title of the Mini Linguistic State Examination (or 'MLSE'). The instrument is intended to be comprehensive yet brief, and capable of being administered by clinicians without special expertise in language assessment.

It would be unrealistic to try to develop versions of the MLSE for all the world's major languages at once, so we will concentrate on two European languages: English and Italian, the native languages of around 360 million and 60 million people, respectively. We will ensure that the same classifications emerge from the MLSE as are currently recognized by expert clinicians by validating the test against expert clinical opinion, and against the patterns of aphasia suffered by patients following a stroke. We will also look for validation in the MR imaging appearances that are known to be associated with each variant. When English and Italian versions of the instrument have been fully developed and validated, we will gather additional data from populations of patients with movement related neurodegenerative disorders (such as Parkinson's disease) in order to gain deeper insights into the effects of these groups of disorders on the brain, and to identify ways in which sufferers' quality of life can be improved through rehabilitation.

Primary progressive aphasia (PPA) is an uncommon clinical manifestation of Alzheimer's, tau- and TDP-43-positive neuropathologies. Three clinical syndromes are recognised: semantic dementia (SD) (progressive loss of word and concept knowledge in association with bilateral anterior temporal atrophy); progressive nonfluent aphasia (PNFA) (distorted speech output with largely preserved comprehension, and cortical atrophy focused on left inferior frontal and insular regions); and logopenic progressive aphasia (LPA) (difficulties with word-finding and sentence repetition, in the presence of atrophy in left posterior cortical regions). The syndromes result from different combinations of dysfunction in fundamental components of the language system: speech fluency, word retrieval, appreciation of word and object meaning, use of grammar, and reading/writing. Syndromes and pathologies are probabilistically related, allowing the latter to be predicted in up to 70% of cases. In light of PPA's relative rarity, increased recognition and coordination of large patient cohorts across multiple sites will be required to study disease biology and clinico-pathological correlations, and enable clinical trials of potential treatments. Progress in both endeavours will benefit from a common method of extracting relevant clinical information from large numbers of individuals. The development of a brief, language specific clinical instrument that can classify patients into one of the three main subtypes of PPA, and to characterise the features of atypical variants, is the central aim of this project. The language battery, based on patterns of impairment across components of language competence, will provide a unified clinical diagnostic standard. Simultaneous development in English and Italian will catalyse and inform the creation of equivalent versions in other major languages of the world, providing a common descriptive clinical currency for the sharing of larger data sets.

The outputs of this project will have multiple users outside the immediate community of neurodegenerative disease researchers. Important beneficiaries and impacts will include:

1) Pharmaceutical and biotechnology industries and clinical trials organizations, for whom the MLSE will provide a fully validated method of assessing outcomes in trials of treatments for either symptom alleviation or disease modification in frontotemporal dementia. The availability of parallel versions of the test in multiple languages will remove a major obstacle to the design of robust clinical trials involving international consortia. Opportunities for engagement with this sector will arise as soon as validation of the instrument has been completed and accepted for publication, and will continue for as long as treatments for the non-Alzheimer dementias remain an active goal of biomedical research. Although the applicants propose that the test should be freely usable for clinical purposes and in the interests of academic research, the IP (ownership of which would be shared among the Institutions involved in its development) would be subject to licensing for commercial use.

2) The use of machine learning methods in test data analysis, will stimulate interest from the computer and software industries, as well as computer science academics working in the healthcare and health informatics field. Individuals from these groups will be keen to engage in data sharing arrangements, and to benefit from the expertise of the investigators and researchers involved in the project. The involvement as project partner of Prof Chesi, who already has a high profile in both academic and industrial sectors, will greatly facilitate the recognition and uptake of collaborative opportunities during the lifetime of the project and within five to ten years after its completion.

3) Globalisation has provided an increasing number of opportunities for cooperation and collaboration with bioscientists working in developing nations, and the momentum that will be created by the need for parallel versions will immediately engage the academic and clinical communities in these countries, allowing them to interact and share information and expertise more widely. This will contribute to achieving the goals of global healthcare, and help to consolidate existing academic collaborations, foster fresh links, and facilitate industry involvement with increasing numbers of healthcare organisations worldwide.

4) The question of the true prevalence of young onset dementia syndromes such as frontotemporal dementia has not been fully and unequivocally resolved, though the problem is known to be more widespread and have greater socioeconomic impacts than had previously been recognized. An instrument that is sensitive to language changes characteristic of the two commonest causes of dementia before age 65, and to progression of mild abnormalities at 12 month follow-up would be a uniquely valuable tool in any investigation of the incidence and prevalence of such conditions. Accurate information would, in turn, inform health service policy, influence the importance given to young onset dementia research by charitable and public sector grant awarding bodies, and contribute over the longer term to an understanding of young onset dementia syndromes by the wider public.

5) Finally, the generic skills imparted to the research staff employed on this project would include not only those relevant to research (i.e. theory and technique, scientific data and project management) but also to the skills involved in collaboration across national boundaries (communication with other linguistic and cultural groups), which the effects of globalisation will render increasingly important to individual and national competitiveness.