Metrifonate for Alzheimer's disease

Treatment of Alzheimer's disease (AD) is one of the great challenges to neuroscience for the 21st century. Despite intense investigation since the late 1970s, AD research has not provided a cure or therapy that permanently arrests the disease. Yet, over the past 10 years, there has been a considerable effort directed toward finding palliative therapies for AD that act through enhancement of cholinergic neuro-transmission. Therapy with cholinesterase inhibitors (CEIs) clearly affects the symptomatic expression of AD, but the practical benefits have been unclear.

Two articles^1,2 in this issue of Neurology present data from clinical trials with the CEI metrifonate for the treatment of AD. Metrifonate has been available for many years as a treatment for schistosomaisis. In the late 1980s, it was suspected that as a CEI, metrifonate might have activity against AD.³ Recognition of its potential led to the current studies of metrifonate. The results provide additional confirmation of the CEI effect. Unfortunately, metrifonate's benefits are no greater than those of other CEIs.

Metrifonate's pharmacology is unique among the CEIs being used or considered for AD. It has a short half-life in plasma (2 hours) but a very long action in the brain because it is an irreversible CEI. It lacks specificity for central versus peripheral cholinesterases. Metrifonate is largely unbound to plasma proteins (<15%) and is not metabolized through the hepatic P450 system. Donepezil,^4,5 tacrine,⁶ and rivastigmine,⁷ another CEI currently awaiting FDA approval, have different pharmacologic profiles. Which is the best profile? The clinical efficacy would suggest that none of these features make a difference across the CEIs. Patient tolerability differs, however.

The current studies used accepted diagnostic criteria and disease severity assessments. Both were placebo-controlled, randomized trials. Dropout rates were low, and as a consequence, an intent-to-treat analytic strategy adequately reflected the drug's effects. Nearly 90% of patients randomized to metrifonate completed the 12-week study¹ and nearly 80% completed the 24-week study.² The retention rates for metrifonate were comparable to donepezil⁸ but superior to tacrine.⁹

After 12 weeks of treatment, there was a nearly 3-point benefit for metrifonate over placebo¹ on the Alzheimer's Disease Assessment Scale (ADAS-cog).¹⁰ This effect was sustained up to 24 weeks.² Comparisons with other CEIs must be undertaken with caution as the drugs were not tested concurrently. With that caveat, metrifonate, tacrine,⁹ donepezil,⁸ and rivastigmine⁷ have produced similar changes on the ADAS-cog: from slightly less than 3 points to about 5 points compared to placebo over 6 months. Furthermore, changes on the clinician's global assessment ratings¹¹ were also similar despite methodologic differences in the studies. The Clinician's Interview-Based Impression of Change (CIBIC) ratings for all the CEIs have been in the range of 0.25 and 0.5 rating points at 6 months. The effect of metrifonate on the Mini-Mental State Examination (MMSE) compared to the placebo group was positive; modest in the 12-week study¹ but not in the 26-week.² Again, the effects of the other agents on the MMSE are comparable to metrifonate's.

There are multiple demonstrations of efficacy of this class of drug in AD. Although the validity of the initial results with tacrine⁹ were questioned, the subsequent publication of trials for donepezil,⁸ now metrifonate,^1,2 and the abstract presentation of rivastigmine⁷ form a broad foundation of support for the reality of the CEI effect in AD. Data from shorter trials of controlled-release physostigmine¹² and galanthamine¹³ add further evidence for a consistent effect of CEIs on cognition and on global assessments. Nonetheless, because metrifonate's efficacy is not substantially greater than the others, questions about the "value" of CEI use in AD remain unresolved for some observers.

Metrifonate did not produce statistically significant effects on measures of activities of daily living. There are three possible explanations for this observation. One is that the drug's effect is too small. Another is that the duration of therapy was too short to observe differences with treatment. A third possibility is that the wrong measurement tool was used. The combination of a relatively small effect of the drugs plus the short period of study probably accounts for the inconsistent effects of CEIs on daily living skills. Unfortunately, functional assessment is an area for which there is no consensus on the best instrument for measuring change in clinical trials. The trials of donepezil,⁸ metrifonate,^1,2 and tacrine⁹ all worked with different assessment tools. It is simply not possible to make claims about similarities or differences of effects on daily functioning among the different CEIs. Although improvement or stabilization in functional effectiveness of an AD patient is a major goal of therapy, measurement of this ability has been difficult. Galasko et al.¹⁴ have devised a new activities of daily living scale specifically for use in clinical trials. Widespread use in a single instrument would allow rough comparisons across studies.

The 26-week metrifonate study provided the first prospective evidence for beneficial effects of a CEI on mood and behavior in a controlled trial. A prior open-label study of tacrine¹⁵ and a secondary analysis of the 30-week tacrine study¹⁶ suggested this finding. A study of the muscarinic agonist xanomeline¹⁷ was the first study of a cholinomimetic drug to observe the effect in a double-blind trial. In one of the current studies,² the Neuropsychiatric Inventory (NPI),¹⁸ a measure of behavioral abnormalities associated with AD, showed benefits of metrifonate therapy compared to placebo for ratings of apathy, dysphoria, and hallucinations. The observations from metrifonate and the other sources^15-17 lead to interesting speculation on the interplay between cognitive effects and alterations in mood and initiative brought about by metrifonate and perhaps other CEIs. Unfortunately, it is not possible to determine whether the cognitive benefits cause improved behavior or vice versa. More must be learned about the noncognitive effects of cholinomimetic drugs. The question has important bearing on understanding of the mechanism of action of these agents.

Most metrifonate-treated patients remained on drug for the entire study. They experienced a low incidence of gastrointestinal side effects. Metrifonate appeared comparable to donepezil in its tolerability. Both can be given in a once per day dosing regimen. At maximum recommended doses, metrifonate and donepezil are both better tolerated than tacrine. The relative ease of use and good safety profile of metrifonate, along with donepezil, removes an important barrier to treatment that often arose with tacrine use.

There are no substantive criticisms of the studies^1,2 reported here. They were conducted using a methodology similar to other CEI pivotal studies. There are obvious deficiencies, however, that are generic for current AD clinical trials. For example, compared with the overall duration of the disease, the period of therapy was short. The two articles do not tell how long to give CEIs. They did not compare metrifonate with already approved agents. The methods for measuring drug effects are not easily understood by those not involved in AD clinical trials.

In the United States, regulatory standards require that two principal outcome measures-an objective measure of cognitive function such as the ADAS-cog and a clinician's global assessment such as the CIBIC-show significant change for a drug to be approved for the treatment of AD. The methodology used in the current studies was constrained by these regulatory requirements. Is there a better way to measure anti-AD drug effects that might have greater ease of interpretation for nonexperts? Under certain circumstances, there probably is. The study by Sano et al.,¹⁹ in which milestones of severe dementia were the outcome measures, is one such example. However, these methods are not applicable to milder AD patients due to the length of the interval between study entry and the appearance of severe dementia. The use of measures like the ADAS-cog and the CIBIC, although divorced from methods usually used in routine practice, represent a compromise. They are broadly applicable, reliable, and valid,^10,11 and therefore reasonably well suited for studies of 6 months' duration with mild-to-moderate AD patients. With more studies similar to those of Morris et al. and Cummings et al., clinicians will gain increasing familiarity with their quantitative aspects. When more potent anti-AD agents appear, neurologists will know what a"bigger" effect is.

An important question for practicing neurologists is whether the effects of metrifonate, tacrine, donepezil, and the others are big enough to justify their widespread use in patients with mild-to-moderate AD. Even if improvement does not occur, their effects in the clinical trials^1,2,7-9 and elsewhere²⁰ seem to translate into a delay in the symptomatic progression of AD for 6 to 12 months. The phase of the disease from diagnosis to severe dementia lasts roughly only 3 to 6 years.²¹ Even assuming that CEIs are strictly palliative and have no effect on the biologic progression of the disease, delay of symptomatic decline of 6 to 12 months is valuable to AD patients. Speculation that cholinesterases play a role in amyloidogenesis^22,23 are not needed to bolster this argument. Given the tolerability of donepezil and metrifonate, one of these drugs should be recommended routinely to patients with mild-to-moderate AD. Furthermore, treatment should be continued until patients reach the stage of severe dementia.

Recommendation of CEIs does not imply that CEIs achieve the goal of definitive therapy in AD. Drug development programs must do better than the CEIs, not equal them. More potent palliative therapies are needed. However, a more fruitful approach to ameliorating the burden of AD may be through preventive therapies. Neuropathologic studies have shown that the morphologic antecedents of AD are present many years before the disease becomes clinically manifest.²⁴ AD has a preclinical prodrome of at least several years' duration.²⁵ Early detection and treatment of the preclinical AD syndrome might offer far better outcomes than treatment of symptomatic mild-to-moderate AD. Clinically apparent AD, however mild, represents "end-stage" disease. The effects of metrifonate and the other CEIs are undoubtedly constrained by the advanced degree of pathology present even in patients with mild, but symptomatic, AD.

Disclosure

Dr. Knopman has received compensation as a consultant to Parke-Davis, Athena Neuroscience, Eisai/Pfizer, Novartis, Wyeth-Ayerst, and Bayer, the sponsor of metrifonate. He is currently participating in clinical trials with Sigma-Tau, Bayer, Forest Laboratories, and Searle.