Metformin in Amnestic Mild Cognitive Impairment

1. SPECIFIC AIMS:

The goal of this study is to conduct a phase II placebo controlled randomized trial of
metformin lasting 12 months in the prevention of memory decline in 80 persons with
amnestic MCI in order to collect preliminary data on efficacy, safety, and feasibility.

- Our primary aim is to compare changes in total recall of the Selective Reminding
Test and changes in the modified Alzheimer's Disease Assessment Scale-Cognitive
Subscale (ADAS-COG) between metformin and placebo on an intent to treat basis..

- Our secondary aim is to compare relative glucose uptake (rCMRgl) in the posterior
cingulate-precuneus measured with brain [18]F-labeled 2-deoxy-2-fluoro-D-glucose
(FDG) positron emission tomography (PET).

- Our exploratory aim is to compare plasma amyloid beta (Aβ) 42 between metformin and
placebo.

2. BACKGROUND AND SIGNIFICANCE:

- Burden of Alzheimer's disease. The prevalence of Alzheimer's disease (AD) is
expected to quadruple by the year 2047. There are no known curative or preventive
measures for AD. Current treatment options for AD only address symptoms, and no
treatments are available that focus on delaying the actual disease process. One of
the currently accepted hypothesis of the pathogenesis of AD is that the main
culprit is the accumulation of Aβ in the brain, and this process has become a
target for treatments and preventive measures. Amnestic mild cognitive impairment
(MCI) has been used to describe a transitional state between normal cognitive
function and AD, and has thus been targeted for interventions. Persons with MCI
progress to AD at the rate of nearly 10% to 15% per year. The criteria most
commonly used for the definition of AD dementia from MCI. We propose to use these
criteria with slight modification to recruit persons for a pilot trail of AD
prevention in persons with amnestic MCI.

- Hyperinsulinemia, diabetes, and risk of AD. Peripheral hyperinsulinemia (high
insulin levels) potentially impair Aβ clearance, and in this study we are proposing
to use metformin, an insulin lowering agent, to prevent AD by improving Aβ
clearance in the brain. The insulin resistance syndrome and hyperinsulinemia are
common in individuals with and without diabetes, and are related to increased risk
of cardiovascular and cerebrovascular outcomes. Hyperinsulinemia predicts the
development of diabetes; therefore, diabetes can be considered a consequence and a
marker of past hyperinsulinemia. According to NHANES III data, more than 40% of the
population over the age of 60 years has problems of glucose intolerance or
diabetes, all related to insulin resistance and hyperinsulinemia. We found that the
risk of AD in individuals without diabetes increases with increasing levels of
fasting insulin, and that high insulin levels are related to a faster decline in
memory scores. The high prevalence of hyperinsulinemia and diabetes (49% of the
elderly in Northern Manhattan) and its biological plausibility as a risk factor for
cognitive decline and AD has attracted increasing attention. In this application we
are targeting hyperinsulinemia, the most important risk factor for AD identified in
the elderly population of Northern Manhattan. The risk of AD attributable to
hyperinsulinemia or diabetes in Northern Manhattan was 39%, and is higher in
Hispanics and African-Americans, who have a higher prevalence of diabetes and
insulin resistance, and will comprise the majority of our sample

- Choice of metformin as treatment agent:. The Diabetes prevention program (DPP) is a
trial of metformin vs. lifestyle intervention (diet and exercise) vs. Placebo in
the prevention of onset of diabetes in 3,234 individuals without diabetes at
baseline. It found that both metformin and lifestyle intervention were successful
in preventing diabetes vs. placebo in 3 years of follow-up and in reducing insulin
levels. Serious adverse events of metformin including hospitalizations and
hypoglycemia were similar to placebo. More recently, another agent that lowers
insulin levels, rosiglitazone, was tested in similar fashion for the prevention of
diabetes and found effective, and is now being tested in AD given its insulin
decreasing properties. However, rosiglitazone's safety profile is not as good as
metformin- it can cause, edema and heart failure(21), and is not widely using in
clinical practice in persons without diabetes, as metformin is.

3. METHODS. PARTICIPANTS: In general the participants are persons who meet criteria for
Amnestic Mild Cognitive Impairment, do not have treated diabetes, and are overweight or
obese (BMI >25 Kg/m2).

- Inclusion criteria:

- Memory complaint expressed by the participant and recognized by the informant.
The memory complaint must represent a change from previous functioning based
on information provided by both subject and informant.

- Age range: 55 years to 90 years. The main rationale for this inclusion
criteria is to follow the standard set by the ADCS

- Sex distribution: men and women. - Languages: fluent in English or Spanish. -
MMSE equal or more than 20. The rationale for this cutoff is that the
population of Washington Heights is heterogeneous in educational and cultural
background. Persons with low educational attainment may have a score of 20
without having dementia.

- Subjects must fulfill criteria for amnestic mild cognitive impairment (MCI)
according to criteria developed for the population of Northern Manhattan by
one of the co-investigators in this study Jennifer Manly. The diagnosis of
mild cognitive impairment is reached by consensus between Drs. Luchsinger and
Dr. Manly based on information from the neuropsychological battery, the
presence of memory complaints, and the clinical dementia rating. This is the
standard procedure used in the Washington Heights Inwood Columbia Aging
Project (WHICAP) for the diagnosis of amnestic MCI. Guidelines for the
diagnosis of MCI: Subjects must score below a predetermined cut-off score on
the logical memory II delayed paragraph recall sub-test of the Wechsler Memory
Scale Revised (WMS-R) or the selective reminding test (SRT). For the logical
memory II delayed paragraph recall the cutoffs are:

1. less than or equal to 8 for 16 or more years of education.

2. less than or equal to 4 for 8-15 years of education;

3. less than or equal to 2 for 0-7 years of education. For the SRT subjects
must score below a predetermined cut-off score. The cutoffs for the SRT
are calculated for each individual based on their age, gender, and
education level. For example, for a 75-year old African American woman
with 6 years of education, an SRT total recall score of 25 yields a
T-score of 42, which is in the normal range, while for an AA woman of the
same age with a master's degree (18 years of school), the T-score is 34
which is in the impaired range. This woman would be classified as MCI if
she met the criteria for function.

- Global CDR score must be 0.5 at screening. The memory box score must be 0.5 or
1.0, with no more than two box scores other than memory rated as high as 1.0
and no box score rated greater than 1.0. - *Subjects without a known history
of diabetes or diabetes that has never been treated with medications. If
diabetes is diagnosed during screening or they have a history of diabetes not
treated in the last 12 months they will be excluded if their HbA1c is > 6.5.
In addition, a diagnosis of diabetes can be made if the HbA1c is 6.5% or more
(American Diabetes Association, January 2010). The reason our study does not
include persons with diabetes is that some medications to treat diabetes
(sulfonylureas, insulin) INCREASE insulin levels. This would interfere with
our study, which is based on the premise that lowering insulin levels with
metformin will decrease the risk of memory impairment. The ADA recommends a
target of < 7 of HbA1c for diabetes treatment. Thus, persons with a HbA1c < 7
with a new diagnosis of diabetes do not require pharmacologic treatment. We
propose to include in our study persons with a new or previous diagnosis of
diabetes, never treated with any drug, who have a HbA1c of 6.9% or less. The
rationale for this cutoff is that we want to diminish the chances that someone
enrolled in the study with new diabetes or old diabetes without pharmacologic
treatment will require treatment during the 12 months of the study. We will
provide all participants with the results of their testing including HbA1c.
The participants and their physicians will have the discretion to start
diabetes treatment. If this happens we will record such information and
continue following the participants as randomized. These participants (those
diagnosed with diabetes started on treatment outside of the study) would still
be considered in Intent-to-treat analyses. All other inclusion and exclusion
criteria still apply to these participants including body mass index.

- Overweight or obese by National Heart, Lung, and Blood Institute (NHLBI)
criteria (BMI of more or equal of 25 kg/ m2)(33). The main rationale for the
use of this criteria is that they predict insulin resistance in the general
population and we need to use criteria that are generalizable and standard for
a phase III trial. Measurement of insulin for the diagnosis of insulin
resistance is not standard in clinical practice, and there are no absolute
definitions of what constitutes insulin resistance. Using BMI criteria as a
surrogate of insulin resistance is an acceptable alternative used in research
and clinical practice. Glucose intolerance is not necessarily a proxy for
insulin resistance and may be a sign of insulin deficiency. Thus, we will not
use glucose intolerance as a criterion for entering the study.

- No contraindications to metformin treatment.

- Hachinski score less or equal to 4.

- Hamilton score less or equal to 12 on the 17 item scale.

- General cognition and functional performance such that a diagnosis of dementia
cannot be made at the time of screening based on DSM-IV criteria.

- Vision and hearing must be sufficient for compliance with testing procedures.

- Exclusion criteria:

- Individuals with dementia

- MMSE < 20 - At the time of telephone contact for screening, we may conduct the
Telephone Interview for Cognitive Status (TICS) if the potential participant
agrees. This procedure is currently followed in other Studies at Columbia
University to screen out persons who are unlikely to have any memory
impairment (we are recruiting persons with mild cognitive impairment). A Score
> 34 on the Telephone Interview for Cognitive Status (TICS) out of 41 is
considered normal cognition. Persons with this score will not be invited to
participate.

- Subjects with neurologic diseases associated to neurologic deficits.

- Subjects with current psychiatric diagnoses such as depression, bipolar
disorder or schizophrenia. - Normal individuals without cognitive complaints.

- Subjects with uncontrolled hypertension (systolic blood pressure more than 160
mmHg or diastolic blood pressure more than 95 mmHg.

- Subjects with a history of active cancer or cancer within last five years,
with the exception of squamous or basal cell carcinoma of the skin.

- Subjects who for any reason may not complete the study as judged by the study
physician.

- Subjects with a known history of diabetes treated with medications.

- Subjects with a new or old diagnosis of diabetes, never treated, with a
hemoglobin A1c of more than 6.5

- Contraindications to metformin: Contraindications to metformin use include a
creatinine of > 1.5, liver disease by history or by elevated transaminases,
congestive heart failure, and alcohol abuse. We will also exclude subjects
with a history of intolerance to metformin. - Use of cholinesterase
inhibitors. The rationale for this exclusion is that the ADCS showed a
beneficial effect of donepezil at 12 months(16) which could confound our
study. We do not have survey data on the use of cholinesterase inhibitors, but
believe that their use is uncommon in persons with AMCI, particularly if they
are recruited from the community.

- Exclusion criteria for brain imaging study:

- Presence of diabetes, even if the HbA1c is less or equal to 6.5. The reason
for this is that FDG PET uptake will be affected by glucose levels.

- Inability to lie down for any reason.

- Presence of any metallic implant.

- Claustrophobia.

- Any contraindication to MRI or FDG PET as determined by the staff of the hatch
center (MRI) or Kreitchman center (PET). A safety screen form from the MRI
research center is attached.

SCREENING AND RECRUITMENT: We will advertise within the medical center, using free
newsletters in the Community of Washington heights, using paid print media, using free and
paid internet media, and using radio. If we are contacted for participation, we will screen
for exclusion criteria. If persons feel comfortable with cognitive testing over the phone, we
will administer the Telephone Interview for Cognitive Status (TICS). We will have access to
WebCis, the internet based clinical information system from Columbia University Medical
Center (CUMC), which we will use to ascertain exclusion criteria. If the person meets
criteria for AMCI, they will be invited to participate in the study and randomized to
metformin or placebo within one month.

The sources of the participants will be:1) The Associates in Internal Medicine (AIM) practice
of Columbia University; 2) the Memory Disorders Clinic (MDC) at AIM; 3) The Community of
Washington Heights. We expect that 50% of participants will be Hispanic, 30% Black, and 20%
Non-Hispanic White, 70% women, and 30% men.

RANDOMIZATION. We propose the use of random permuted blocks as the method of randomization to
ensure balance in the groups given the small sample size. If participants are included in the
study we will do APOE-ε4 genotyping and use this information for block randomization.

SCHEDULE OF ASSESSMENTS. We will collect medical history and concomitant medication data at
every visit. Participants will be examined at 3 month intervals. Imaging studies will be
conducted only at baseline and after the 12 month visit.

METFORMIN AND PLACEBO. Metformin Dosing Schedule and side effects: Metformin and matching
placebo will be provided by Merck-Lipha of France. The medication and placebo will be managed
and dispensed by the Research Pharmacy at Columbia University Medical Center. Persons will
first be given metformin 500 mg once a day at baseline. At day-7 the dose will be increased
to 500 mg twice a day. At day 14, the dose will be increased to 500 mg three times a day. At
day 21, the dose will be recommended to 1000 mg twice a day. At day 28, we will call the
patient via telephone to ascertain that they are tolerating the dose of 1000 mg twice a day,
the usual maximum dose recommended in clinical practice. If the participants report not
tolerating a dose of metformin, they will be asked to remain at the lowest tolerated dose.
The most common side effect of metformin is gastrointestinal intolerance. At baseline, 3
month, 6 month, 9 month visit and at the end of the study we will measure renal function,
liver function, and complete blood count to monitor side effects. This information will be
made available to the data safety monitoring board of the study.

NEUROPSYCHOLOGICAL BATTERY. All measures will be available in English and Spanish. The
neuropsychological battery includes our 2 co-primary outcomes, the Buschke Selective
Reminding test, and the modified ADAS-Cog. In addition, our battery includes the ADCS
Clinical Global Impression of Change-Mild cognitive impairment, the Clinical Dementia Rating,
the Digit Span Backwards, the Hamilton Rating Scale for Depression, the Logical Memory Test,
the Mini Mental State Examination, and the neuropsychiatric inventory.

OTHER ASSESSMENTS:

We will assess body mass index (BMI), blood pressure, heart rate, respiratory rate, and
conduct a physical exam to search for contraindications to metformin including cardiac
disease. Will also perform an EKG at baseline.

LABORATORY MEASURES. Plasma Aβ40 and Aβ42 have been suggested as markers of AD and high Aβ42
has been shown to predict AD progression in Northern Manhattan . They will be measured in all
participants at baseline and at the end of the study from EDTA plasma stored a -80 degrees F
using published procedures.

General chemistry, liver function tests, and complete blood count will be done as shown in
the timetable to monitor for contraindications and side effects of metformin. In persons with
no available information on TSH, B12, and RPR, necessary measures to rule out secondary
cognitive impairment, we will obtain them at baseline. HsCRP, insulin, lipids, adiponectin,
and HbA1c are inflammatory and metabolic intermediate variables and will be measured to
ascertain the effects of metformin and to explore mechanisms for a benefit of metformin on
the outcomes. APOE-ε4 genotyping will also be conducted.

BRAIN IMAGING. A subsample of the study will compare on an intention-to-treat basis between
the metformin and placebo group mean changes from the beginning to end of the trial in uptake
of fluorodeoxyglucose (FDG) in the posterior cingulate-precuneous measured with brain
positron emission tomography (PET) in 40 non-diabetic patients out of the 80 study subjects.
PET imaging will be performed with a HR+ scanner at Columbia Kreitchman PET Center. All image
analysis will be performed with MEDX and FSL (FMRIB) Software Library). There will also be a
brain magnetic resonance imaging (MRI). MRI images co-registered to the PET images allow
accurate structural localization of metabolic changes and correction of brain atrophy.
Participants will be scanned on a 1.5T Philips Intera dedicated research scanner. Four sets
of structural images will be obtained from each subject during their MRI scanning session.
The total imaging time will be 30 minutes. As part of standard procedures, a neuroradiologist
conducts a clinical evaluation to rule out tumors, strokes, and other lesions. Any
significant abnormality is discussed with the study investigator, who then takes appropriate
action, including discussion of the abnormality with the participant, and with the
participant's permission, with the participant's physician of choice. If the participant has
no physician, we will facilitate one at Columbia University Medical Center. PET and MR images
will be spatially normalized to a custom template image in standard Montreal Neurological
Institute (MNI) space. The spatial normalization will allow assessing any changes in the
metabolic rate of glucose as a result of the intervention through a statistical parametric
mapping approach. These spatial parameters will be applied to the co-registered PET images
thereby transforming them into standardized space.

STATISTICAL ANALYSES. The primary analyses will be conducted on an intention-to-treat basis.
We will use ANACOVA to compare changes from baseline to 12 months in all outcomes between
metformin and placebo. We will adjust for variables found to be different between the
metformin and placebo groups at baseline. The primary analyses are those for the primary aim,
secondary aim, and exploratory aim.

We will also secondary analyses examining completers, and also examining changes in metabolic
markers and other cognitive measures.

I

Inclusion Criteria:

- Age range: 55 to 90 years;

- Sex distribution: men and women;

- Languages: fluent in English or Spanish.

- Subjects must score below a pre-determined cut-off score on the logical memory II
delayed paragraph recall sub-test of the Wechsler Memory Scale Revised (WMS-R)26:

1. less than or equal to 8 for 16 or more years of education.;

2. less than or equal to 4 for 8-15 years of education;

3. less than or equal to 2 for 0-7 years of education.

- Global Clinical dementia rating (CDR) score must be 0.5 at screening. The memory box
score must be 0.5 or 1.0, with no more than two box scores other than memory rated as
high as 1.0 and no box score rated greater than 1.0.

- Subjects without a known history of diabetes or diabetes that has never been treated
with medications. If diabetes is diagnosed during screening or they have a history of
diabetes not treated in the last 12 months they will be excluded if their HbA1c is >
6.5.

- BMI ≥ 25 kg/m2

- No contraindications to metformin treatment.

- General cognition and functional performance such that a diagnosis of dementia cannot
be made at the time of screening based on DSM-IV criteria.

- Vision and hearing must be sufficient for compliance with testing procedures.

Exclusion Criteria:

- Subjects with uncontrolled hypertension (systolic blood pressure ≥ 160 mmHg or
diastolic blood pressure ≥ 95 mmHg;

- Subjects with a history of active cancer or cancer within last five years, with the
exception of squamous or basal cell carcinoma of the skin;

- Subjects who for any reason may not complete the study as judged by the study
physician;

- Abnormal TSH, B12, and RPR.

- Contraindications to metformin use include a creatinine of > 1.5, liver disease by
history or by elevated transaminases, congestive heart failure, and alcohol abuse.

- We will also exclude subjects with a history of intolerance to metformin.