A Preliminary Study of d-Cycloserine Treatment in Alzheimer's Disease

The prevalence of dementia in the United States has been estimated to be 2 to 4 million patients. Of these, 50% to 60% of cases are of the Alzheimer's type (AD).¹ Although the cholinergic deficits in the central nervous system are considered to be of primary importance in AD, a number of other neurotransmitters are severely affected, including the excitatory amino acids (EAAs), aspartate, and glutamate.² The loss of both presynaptic EAAs and their postsynaptic receptors suggests that glutamatergic terminal degeneration and deficient EAA neurotransmission may be related to the symptoms of AD. Evidence for these changes includes selective decreases in CSF concentrations of EAAs, reduced d-aspartate uptake,³ and decreased numbers of N-methyl-d-aspartate (NMDA) receptors in the frontal cortex and hippocampus in AD subjects.⁴

The suggestion of a therapeutic role for EAA agonists in treating cognitive deficits of AD comes from several sources. The NMDA subtype of glutamate receptors is involved in neuronal activities underlying memory. NMDA antagonists impair long-term potentiation in the hippocampus, a cellular model of memory.⁵ Conversely, drugs that potentiate NMDA receptors facilitate learning in animals.⁶ Thus, use of EAA agonists to enhance glutamatergic transmission has been proposed as a treatment for AD. However, excessive stimulation of NMDA receptors is neurotoxic, and the potential degenerative effects mediated by EAAs and other agonists lead to the speculation that they would not be good candidates. The NMDA receptor is a complex molecule with multiple modulation sites. The manipulation of the glycine site of the NMDA receptors appears to be a safer way to enhance the activity in the remaining cortical neurons in AD.

d-cycloserine, a broad-spectrum antibiotic used to treat tuberculosis at high doses (500–1,000 mg/day), exhibits partial agonist activity on NMDA glycine sites. It readily crosses the blood–brain barrier. At low doses, d-cycloserine facilitated activation of NMDA receptors in AD brains⁷ and enhanced learning in animals through the same mechanism.⁸ In addition, a single dose of d-cycloserine significantly improved cognition in scopolamine-induced cognitive symptoms in human subjects.⁹ A dose of 15 mg was superior to doses of 5 mg and 50 mg. We conducted a pilot double-blind, crossover study of 15 mg daily dose of d-cycloserine to assess its efficacy and safety in treating the cognitive deficits of patients with AD.

METHODS

After a consensus diagnosis was established by a research psychiatrist, a neurologist, and a neuropsychologist, 10 patients (6 women and 4 men) who met NINCDS-ADRDA criteria for probable AD¹⁰ were recruited into the study. All subjects' Hachinski Ischemia scores were less than 4.¹¹ Both they and their caregivers provided informed consent prior to participation and after the procedure had been fully explained. Their mean ages (±SD) were 74.7±8.5 years, and their symptoms had been present for 5.7±2.1 years. Their Mini-Mental State Examination (MMSE) scores ranged between 10 and 25. All patients were randomly assigned, under double-blind conditions, to receive either a 4-week trial of 15 mg d-cycloserine per day followed by 4 weeks of placebo, or the reverse order of treatment. The choice of a 4-week trial period was based on our study of schizophrenic patients who improved their cognitive symptoms after a minimum of 2 weeks of treatment with d-cycloserine.¹²

Evaluations included the MMSE, Alzheimer's Disease Assessment Scale (ADAS), Clinical Global Impression of Change (CGIC), Instrumental Activities of Daily Living (IADL), Physical Self Maintenance scales (PSM), and Blessed Activities of Daily Living Scale (Blessed). These were administered by a trained psychometrician at baseline and at the end of each 4-week period. At each visit, both the patient and caregiver were systematically asked if they experienced or observed any side effect. In addition, physical and neurological examinations were performed at each visit to look for adverse effects. Each subject had a caregiver who completed necessary evaluations and supervised medication compliance.

The efficacy of d-cycloserine was tested by analysis of variance (ANOVA ) and nonparametric Mann-Whitney U-test. The ANOVA was also used to assess the probability of carryover effects.

RESULTS

No side effects with d-cycloserine treatment were noted by patients, caregivers, or the research psychiatrists. However, using one-way ANOVA, we could find no therapeutic effect on any outcome measure (Table 1). For the patients who received d-cycloserine first, no carryover effect was noted by ANOVA. Furthermore, since the elimination half-life of d-cycloserine is about 12 hours,¹³ it is unlikely to have a residual effect 4 weeks after discontinuation. When the two arms of the crossover study were compared, the results of the cognitive tests were not different (data not presented), arguing against a practice effect on the cognitive tests. The nonparametric Mann-Whitney U-tests of ADAS (U=48, P=0.71), CGIC (41.5, 0.43), MMSE (49, 0.94), IADL (48, 0.88), PSM (36, 0.27), and Blessed (36, 0.3) did not reveal any significant change after the 4-week trial of d-cycloserine when compared with the placebo phase, either.

DISCUSSION

None of the assessment instruments demonstrated clinical benefit of d-cycloserine at the 15-mg dose. However, the choice of this dosage was based on the studies of scopolamine-induced cognitive deficits that were improved by 15 mg of d-cycloserine.⁹ Furthermore, d-cycloserine appeared to have a narrow therapeutic range for improving cognition,¹² a finding supported by its action as a partial agonist of the glycine site of the NMDA receptor. Nonetheless, a higher dose may be required to enhance cognition in AD patients. The negative result may be also due to the small number of subjects and consequent lack of statistical power to detect a small magnitude of improvement in cognition of AD patients.

The stable outcome measures across three phases of the study argue against a rapidly degenerative process during the study period. However, the issues need to be clarified by placebo-controlled parallel group study. AD patients likely have varying degrees of loss of glutamatergic neurotransmission, and d-cycloserine may be beneficial only at the early stage of AD when the neuronal architecture is still intact. Improvement of cognitive function by augmenting glutamatergic neurotransmission in AD may require long-term treatment to “reestablish” the learning circuitry.

While we were conducting the study, another two series of AD patients receiving d-cycloserine were reported.^14,15 In the report of Randolph et al.,¹⁵ although no therapeutic effect was demonstrated, 7 of 12 patients had ADAS cognitive subtest scores improve by at least 4 points. In the report of Fakouhi et al.,¹⁴ patients receiving 30 mg d-cycloserine a day improved significantly on the priming effect of implicit memory but not on the other cognitive tasks. Taking these reports together with our findings, we conclude that it is premature to decide on the effectiveness of d-cycloserine in AD. Because of the substantial loss of glutamatergic neurotransmission in AD and the scarcity of pharmacotherapeutic agents, it is important to pursue higher doses of d-cycloserine or other glutamatergic agents that can safely enhance cognition in AD.

ACKNOWLEDGMENTS

This work was supported in part by American Federation of Aging Research, a NARSAD Young Investigator Award, a Stanley Foundation Research Award, the Canavan Foundation, and the Peter and Elizabeth C. Tower Foundation (G.T.).

TABLE 1.

Enlarge table