A BRIEF HISTORY OF THE WORK

To be at its most successful, clinical research must run hand-in-hand with provision of good quality care, relevant to the patient's needs: the one enhances the other. With this objective, the Therapeutics Research Group was set up, in 1984, under the aegis of Sir Christopher Booth, Director of the MRC's Clinical Research Centre (CRC) at Northwick Park. Its initial remit was to study problems relevant to the aged population served by a District General Hospital. Since the closure of CRC in 1993, the research has focused on the role of occult inflammation in the aetiology and pathogenesis of neuropsychiatric disease, finding an ideal postgraduate academic base in Clinical Neuropharmacology at the Institute of Psychiatry, with research clinics in the Maudsley. A Steering Committee of eminent scientists and clinicians has been formed. In July 2002, it held its inaugural meeting with the multidisciplinary team of principal investigators from King's College London, Imperial College and University College, School of Hygiene and Tropical Medicine and collaborators from the Central Public Health Laboratory Service in North London.
The tight efficacy study, with outcome data of which the basic pharmacologist would be proud and economy in sample size, is a hallmark of the Therapeutics Research Group's work. Breaking down clinical and academic boundaries has been crucial to problem solving. The current programme builds on wide ranging experience in clinical pharmacology and therapeutics in the entire age spectrum. Work on the aetiology of parkinsonism grew from developing methodology to answer fundamental therapeutic questions, such as tolerance to anti-parkinsonian treatment and the existence, and drug responsiveness, of cognitive inefficiency ("bradyphrenia").
Definition of pre-presentation parkinsonism in functional terms, and the search for biomarkers, have been cornerstones of the work. Homing in on what's driving it is the motivation and excitement, intervention in a pre-clinical state the goal. Longitudinal, objective follow-up is a must. Conventional approaches may have limited potential: major advances are not made on bandwagons, but by thinking the unthinkable, not accepting dogma, and reviewing the obvious.

PART 1: NOVEL METHODS OF MEASURING THE SYNDROME

The applicability of the objective assessments described here goes far beyond its original role in the study of ageing and parkinsonism. That homogeneity does not come with age, is as apparent in pharmacodynamic as in kinetic studies.
In parkinsonism, the global score holds sway, but is a blunt instrument. Simple relevant tests of performance are advocated, but often with inadequate attention to their sensitivity, specificity and reliability. The time to walk an individually set distance, return to and sit in a chair, and the rate of progress at fastening a set of buttons are reliable, but lack sensitivity to dopaminergic effects. In contrast, gait analysis at free walking speed has excellent reliability, and usefully defines treatment effects.
Invention, design and development of equipment are tailored to needs. Devices should be inexpensive, not require a specialist operator and useable in the clinical (including domestic) context. Real time and storage telemetry is used to allow monitoring during everyday activities. Equipment is designed to be small and unobtrusive, and thus acceptable and suited to the frail and disabled. Output is designed to be readily interpreted by the non-specialist. This facilitates feedback to the patient. The plan is to bring precise, objective measurement to routine practice, as well as clinical research. A secondary, but important consideration, is commercialization (e.g. every physiotherapist should have a "pocket gait analysis laboratory": it should be a tool of the trade).
Targets for objective assessment of parkinsonism are the cardinal signs, major associated disability (e.g. cognitive inefficiency, inability to turn in bed) and events (sleep apnoea, falls). Normality often seems to drift into abnormality. The more one studies parkinsonism, Alzheimer's disease, Lewy Body dementia and depression, the more blurred the clinical margins. Aetiological clues may lie in the overlap: divisions imposed by man may be a diversion to fundamental research. Continuous physical and cognitive measures are needed, cross-referenced to the usual clinical diagnostic threshold. Reliable longitudinal follow-up is essential.

Figure 1.

In many in vitro pharmacological experiments, response would not be so obvious from the raw trace as shown here. The amplitude of the waveform represents swing length, the flattened peaks and troughs the hesitancy between swings. This is the response to Madopar CR (controlled-release levodopa/benserazide) at ½ and 2½ h post-treatment. The lower trace gives foot separation at mid-swing. Resting separation does not distinguish the two time points, but the patient was walking with a dangerously narrow base at ½ h, before her medicine took effect.

Delay in answering questions, and not being able to follow a story line, embarrasses people with parkinsonism and their carers. Is the "bradyphenia" associated with parkinsonism a nosological entity, or can it be accounted for by depression or dementia, or exogenous substances? Naville, 1922, described a slowing of cognitive processing associated with parkinsonism, consequent on pandemic encephalitis lethargica. Wilson, 1947, further described this lethargy of mind as "distinguished by a lack of interest, initiative, attention, concentration; by fatigue and slow reactions: uncommunicative and wishing to be left alone, the patient sits and does nothing unless exhorted to rouse himself". However, half a centuary later it was still controversial: a simple model was needed. The amount of quickening in reaction time in response to a warning is, surely, a reflection of the efficiency in processing that warning. We found that central processing, as measured by this quickening, had a component which could be explained by the presence of idiopathic parkinsonism, but not by age, mental test and depression scores, or the consumption of exogenous substances. The discriminant ability for parkinsonism was characterised by high specificity, but low sensitivity. We went on to examine the slowing of mental processing in response to increased cognitive complexity, in this case achieved by delayed auditory feed back: this was complimentary in discriminating between normality and parkinsonism. The way is now open to mapping the pattern of responsiveness to medicinal interventions, using prospective placebo-controlled protocols.

STRATEGY AND APPROACH
The overall strategy grew from our work at the Medical Research Council's Clinical Research Centre, Harrow, to set the discipline of therapeutics in older people on a firm scientific footing. There are long lists of adverse reactions to medicines, and remarkable statistics on their frequency, but this is not the object of the exercise. There is insufficient solid information on the efficacy of therapeutic interventions.
Throwing large subject numbers at a problem may impress by a high degree of statistical significance despite a small magnitude of effect. Face value endpoints may undermine rational intervention, unless the mode (and, ideally, the mechanism) of action is elucidated. In the old, confounding events and (medical and socio-economic) influences are many, such that "intention-to-treat" analysis may show little benefit, even for undoubtedly efficacious interventions. With better-focused outcome criteria, the whole time consuming procedure of a therapeutic intervention (environmental, medicinal or remedial) finding its proper place in the care of older people could be speeded up.
Screening in old age also needs to be focused. Age-related disorders, and the ageing process itself, need to be examined: how much is reversible and what is preventable by identifying adverse environmental influences? Effort put into defining pre-clinical and covert disease states, and intervention in these, may save resources that would have to be spent at a later date in established disabilities.
The approach was to shift emphasis from arbitrary expectations of the benefit to objective measures of outcome, suitable for and relevant to older people. The outcome of studies in even the frail, aged patient can be as lucid as any in vitro pharmacological experiment. Emphasis on suitable methodology opens up the field to conducting studies with a high statistical power, using small numbers of subjects; to much needed longitudinal studies; and to detection of, and intervention in, early disease states. Our non-invasive assessment of cardinal signs, disability, and events has depended on integrating the discipline of biomedical engineering.

We now apply the strategy and approach, developed at the Clinical Research Centre for therapeutics in the elderly, to the study of neuropsychiatric disease, irrespective of age. The subtle pre-presentation state of parkinsonism may never come to diagnosis, but undoubtedly limits personal potential and career. Frank disease of early onset does not just limit, but can shatter, career prospects. Familial and peer group clusters are a source of concern, but may hold invaluable clues to causation. The routine threshold for diagnosis is an unnecessarily blunt instrument in their definition.
Investigating interactions of environment, ageing and genetic predisposition (Figure 1) is complicated: a pragmatic (hypothesis testing) statistical approach needs to be preceded by an exploratory (hypothesis generating) one using statistical modelling. Incorporating the professional statistical/mathematical expertise of André Charlet also underpins the design/analysis of clinical trials, where effects are likely to be small, confounding influences important, and attrition in sample size and missing data points unavoidable.

Figure 2. Factors involved in the causation of idiopathic parkinsonism.

PART 2: TAILORING TREATMENT

In efficacy studies, objective outcome criteria can give high power with economy in patient numbers, provided there is a direct link between intervention and measured effect. Time sequence effects may mean that measures associated with long term prognosis in younger patients are not useful in the shorter term, and of questionable relevance to the old. Shifting from remote arbitrary expectations to what is desirable in terms of performance, or frequency of events, may provide a firmer basis for exploring efficacy and mechanism. Below are just two examples of what precise outcome measures can contribute to lucidity and economy.

Selegiline was developed in 1964, as a monoaminoxidase-inhibitor antidepressant. Clinical interest only really took off over a quarter of a century later, with laboratory evidence that it could be prophylactic in parkinsonism. As first-line therapy, selegiline had benefit on face-value endpoints (time to requiring dopamine replacement therapy using levodopa, and in full employment) in a double-blind, placebo-controlled, parallel group, multi-centre study of 800 patients.

Whether these coincide with an effect on a cardinal sign(s), or can be explained by cognitive effects, was unclear. Our double-blind, placebo-controlled parallel group study in newly diagnosed, untreated elderly sufferers suggests the global benefit could be psychostimulant. Arousal increased to 6 months, psychomotor activity improved, and tolerance then developed. However, we also measured the cardinal signs: there was a highly significant effect of nature of treatment with respect to time on rigidity, measured as the mean work required displacing the forearm.

It improved gradually with time, rigidity decreasing by, on average, 1.3% per week (Figure 2). The apparent worsening on placebo was not significant. The time course of the improvement in rigidity was not typical of a dopaminergic treatment effect (i.e. an exponential growth occurring mainly in the first week). It could represent neuronal rescue. The sample size needed to elucidate the nature of the effect was 3% of that used in the multi-centre study.

Figure 3. Comparison of mean (95% C.I.) time course in rigidity, expressed as ratio of mean work required displacing the forearm during treatment to that pre-treatment, on active (¾) and placebo (---) selegiline.

There had been much discussion about tolerance (declining efficacy with time) to levodopa, but no objective evidence for or against it. We took just 8 patients with end of dose wearing-off effect, receiving conventional levodopa/decarboxylase inhibitor preparations, and measured stride length, in relation to single dose challenges with placebo, and one and two tablets of a controlled release levodopa/carbidopa, given according to a double-blind, randomised, cross-over design. After 5 months on maintenance therapy with the controlled release, the single dose challenges were repeated. The placebo challenges showed a clinically useful improvement in background performance, as measured by stride length, during maintenance therapy with the controlled release. The response to two tablets had not varied significantly between the two time points, and was not significantly different from that for the initial challenge with one tablet.

However, the response to the second challenge with one tablet did differed significantly from that of the first: the response had become submaximal. There appeared to be tolerance to an acute challenge at the lower dose. It could be explained, in part, by the accumulation of a long half-time metabolite. The size of the tolerance effect was such that this objective study with only 8 patients had a very high (98%) power to detect it. Of the clinical ratings used, tremor (scale 0 to 4) was the most sensitive. However, had tremor been our primary outcome criterion, a five fold increment in the number of patient recruits would have been required to achieve the same power.

Part 3 will address "homing in on what is driving it"

Call for Volunteers

If you have been diagnosed as having Parkinsonism, but have not yet been exposed to anti-parkinsonian medication, you can contribute enormously to understanding the driving force behind the illness, and defining the optimal strategy for arresting it.

Further information from: Drs Sylvia & John Dobbs,
Tel 01442 873571, fax 0208 868 3350
Email dobbs@wellers.demon.co.uk giving subject as PD research enquiry.