Page 1

Disease progression model

in subjects with mild

cognitive impairment from

the Alzheimer’s disease

neuroimaging initiative:

CSF biomarkers predict

population subtypes

Mahesh N. Samtani,1Nandini Raghavan,1Yingqi Shi,1

Gerald Novak,2Michael Farnum,3Victor Lobanov,3Tim Schultz,3

Eric Yang,3Allitia DiBernardo,2Vaibhav A. Narayan2& the

Alzheimer’ s Disease Neuroimaging Initiative*

Johnson & Johnson Pharmaceutical Research & Development,1Raritan,New Jersey,2Titusville,New

Jersey and3Spring House,Pennsylvania,USA

Correspondence

Dr Mahesh N.Samtani PhD,Johnson &

Johnson Pharmaceutical R&D,Clinical

Pharmacology Department,Advanced

PK/PD Modeling and Simulation

Department,920 Route 202,PRD 2723,

Raritan,NJ 08869,USA.

Tel.:+1 908 704 5367

Fax:+1 908 927 2573

E-mail:msamtani@its.jnj.com

----------------------------------------------------------------------

*Data used in preparation of this article

were obtained from the ADNI database

(http://www.loni.ucla.edu/ADNI).As such,

the investigators within the ADNI

contributed to the design and

implementation of ADNI and/or provided

data but did not participate in analysis or

writing of this report.A complete listing

of ADNI investigators can be found at:

http://adni.loni.ucla.edu/wp-content/

uploads/how_to_apply/ADNI_

Authorship_List.pdf

----------------------------------------------------------------------

Keywords

ADAS-cog,CSF biomarkers,disease

progression,MCI, NONMEM®

----------------------------------------------------------------------

Received

21 June 2011

Accepted

18 April 2012

Accepted Article

Published Online

25 April 2012

WHAT IS ALREADY KNOWN ABOUT

THIS SUBJECT

• Amnestic mild cognitive impairment MCI)

represents the prodromal stage of

Alzheimer’s dementia and this disease

progresses in a non-linear fashion.

• Disease progression depends on a variety of

demographic,biochemical,genetic and

cognitive factors.

WHAT THIS STUDY ADDS

• Baseline CSF biomarkers carry information

about disease pathology and critical

thresholds for these markers (Ab and

p-tau181P) have been identified that allow

segregation of the population into MCI

progressers and non-progressers.

AIM

The objective is to develop a semi-mechanistic disease progression

model for mild cognitive impairment (MCI) subjects.The model aims to

describe the longitudinal progression of ADAS-cog scores from the

Alzheimer’s disease neuroimaging initiative trial that had data from

198 MCI subjects with cerebrospinal fluid (CSF) information who were

followed for 3 years.

METHOD

Various covariates were tested on disease progression parameters and

these variables fell into six categories:imaging volumetrics,

biochemical,genetic,demographic,cognitive tests and CSF biomarkers.

RESULTS

CSF biomarkers were associated with both baseline disease score and

disease progression rate in subjects with MCI.Baseline disease score

was also correlated with atrophy measured using hippocampal volume.

Progression rate was also predicted by executive functioning as

measured by the Trail B-test.

CONCLUSION

CSF biomarkers have the ability to discriminate MCI subjects into

sub-populations that exhibit markedly different rates of disease

progression on the ADAS-cog scale.These biomarkers can therefore be

utilized for designing clinical trials enriched with subjects that carry the

underlying disease pathology.

British Journal of Clinical

Pharmacology

DOI:10.1111/j.1365-2125.2012.04308.x

146/Br J Clin Pharmacol/

75:1/146–161

© 2012 Janssen Pharmaceuticals, Inc

British Journal of Clinical Pharmacology © 2012 The British Pharmacological Society

Page 2

Introduction

It is believed that by the time Alzheimer’s disease (AD) is

diagnosed, sufficient neuronal injury has occurred that

reversal of the disease is perhaps unlikely [1]. This has

therefore raised considerable interest in the prodromal

stage of AD involving subjects with mild cognitive impair-

ment(MCI)whoareinthepre-dementiastageofcognitive

dysfunction and therefore could be targeted for therapies

thatcouldpotentiallyprovidebeneficialeffects.Thepreva-

lence rate for MCI around the world is in the range of

14–18%inindividuals?70 yearsofage[2].Inclinicaltrials

andepidemiologicstudiestheannualrateofconversionof

MCIsubjectstodementiaisintherangeof6–15%,whichis

much higher than the incidence rate of dementia of 1–2%

seeninthegeneralpopulation[2].MCIrepresentsaninter-

mediate state of cognitive impairment that is greater than

the level expected for a subject’s education level and age

[3] but does not meet criteria for dementia and does not

compromise activities of daily living.The diagnosis of MCI

is characterized by heterogeneity,varying severity and the

inability to predict disease progression i.e.not all MCI sub-

jects have underlying AD neuropathology [2]. Indeed, not

all cases of MCI progress to AD and a small fraction of

subjects revert back to normal status.However,the clinical

phenotype of amnestic MCI, in which only the domain of

memory is affected, is thought to be degenerative in

nature and these subjects have a high probability of pro-

gression to AD [1].

Neuropsychological assessments are a key component

of detecting and tracking disease progression in clinical

trials because they provide standardized evaluation of

memory and cognitive impairments which are central fea-

tures of MCI.The cognitive component of the AD Assess-

ment Scale (ADAS-cog) has been utilized in the majority of

large scale pharmacologic and naturalistic studies of MCI.

During the past decade several MCI clinical trials have

testedthecurrentpharmacologicagentsusedinthetreat-

ment of AD.None of these clinical trials has achieved their

expected therapeutic end points and therefore there are

no approved treatments for MCI [2]. These results have

caused some concerns about the insufficient sensitivity of

the ADAS-cog scale in mapping and tracking the early

stages of the disease [4].

FromthepublishedCSFbiomarkerstotaltau,phospho-

rylated tau at the threonine 181 position (p-tau181p), and

CSF amyloid beta 1 to 42 peptide (Ab1–42) are considered

as promising markers for inclusion in clinical trials and in

the revised AD diagnostic criteria [5,6].The utility of these

biomarkers is further supported by the newly released

National Institute on Aging/Alzheimer’s Association Diag-

nostic Guidelines for AD that recommend inclusion of

these specific markers for use in research settings, includ-

ing MCI clinical trials [7].Recent reports from the AD neu-

roimaging initiative (ADNI) trial have shown that MCI

subjectsexhibitbimodaldistributionswithrespecttotheir

baseline concentrations of Ab1–42 and p-tau181P[5].These

biomarkers (low CSF Ab1–42 and high p-tau181P) are

thought to reflect the pathologic features associated with

AD.CSF biomarkers have the potential to provide informa-

tion about the probability of disease progression to AD for

an individual MCI patient and the likelihood that this pro-

gression will occur within a defined period.The CSF biom-

arkers considered in the current analysis represent a small

subset of the large number of related biomarkers [6].

However, the CSF biomarkers have a fairly large body of

literature evidence in MCI [8–12] and are therefore consid-

ered as a reasonable starting point. CSF biomarkers were

captured in only 50% of subjects in the ADNI trial and the

currentanalysiswillfocusononlythoseADNIMCIsubjects

who have baseline CSF data available for Ab1–42 and tau

proteins.

Adiseaseprogressionmodelwaspreviouslydeveloped

for patients with AD [13,14] and since the MCI population

represents a distinctly different sub-group in terms of

biomarker characteristics and rates of cognitive deteriora-

tion [5],the current analysis focuses on this earlier stage of

the disease. The recently developed semi-mechanistic

non-linear AD disease progression model was built to (a)

capture the longitudinal change of ADAS-cog scores and

(b)describetherateofprogressionandbaselineADAS-cog

as a function of influential covariates in AD patients [13,

14].In the model, baseline ADAS-cog was associated with

yearssincedementiaonset,hippocampalvolumeandven-

tricular volume. Disease progression rate was dependent

on age, total serum cholesterol, APOE e4 (APOE4) geno-

type,Trail B test,as well as current impairment status mea-

sured by ADAS-cog. Rate of progression was slower for

mildandsevereADpatientsvs.moderateADpatientswho

exhibited a faster rate of disease worsening. One of the

objectives of the current analysis is to assess the applica-

bilityofthisADmodelanditscovariaterelationshipstothe

MCI population.In addition,this analysis incorporates CSF

biomarkers known to characterize MCI subjects with AD

pathology. The combination of total tau concentrations

and the p-tau181p: Ab1–42 ratio predicts the categorical

endpoint of conversion to AD with relatively good sensi-

tivity and specificity [5, 10, 11, 15]. The current analysis

focuses on continuous measures of disease progression

suchasADAS-cogratherthanthecommonlyreportedcat-

egorical end points such as conversion or time to conver-

sion. The emphasis of this analysis is on the mixing

distribution for ADAS-cog change and that for baseline

CSF biomarkers in MCI subjects.The objective is to assess

the degree of correlation between rate of disease progres-

sion as measured by a continuous scale such as ADAS-cog

and baseline CSF biomarker status.This information could

be utilized to enrich clinical trials and may thus enable

successful clinical trials in MCI subjects.The availability of

richlysampledlongtermnaturalisticMCIprogressiondata

from the ADNI public database (available at https://

www.loni.ucla.edu/ADNI) allows assessment of (a) variabil-

Disease progression model in MCI subjects

Br J Clin Pharmacol/

75:1/147

Page 3

ity in this disease state, (b) potential covariates affecting

MCI progression and (c) the ability of ADAS-cog to track

disease progression during the MCI stage.We thus aimed

at developing a non-linear mixed effects model with cova-

riates, incorporating neuropsychological assessments and

structural or chemical biomarkers to describe disease pro-

gression in ADNI MCI subjects.

Methods

Study details

Data used in the preparation of this article were obtained

from ADNI database (http://www.loni.ucla.edu/ADNI); for

up-to-date information, see http://www.adni-info.org.All

ADNI subjects had clinical/neuropsychological assess-

ments and 1.5T MRI measurements, while CSF measure-

ments were performed in only 50% of subjects. MCI

subjects were assessed at 0, 6, 12, 18, 24 and 36 months,

while AD subjects were assessed at 0, 6, 12 and

24 months. ADNI allows public access to all accumulating

data. The dataset available on November 9 2010 (http://

www.loni.ucla.edu/ADNI) was utilized in the current

analysis. This recent download of the database contains

1036 ADAS-cog measurements from 198 MCI subjects

with baseline CSF data. 42.4% of these MCI subjects have

converted to AD at the time of the data download. Other

plasma biomarkers of Ab pathology were not assessed in

the current analysis. A recent report based on the ADNI

data shows that plasma Ab shows mild correlation

with other biomarkers of Ab pathology and is rather

insensitive because health conditions other than AD are

also associated with altered concentrations of plasma Ab

[16]. More importantly, plasma Ab has limited value for

disease classification and modest value as a prognostic

factor for clinical progression [16] and is not considered

further.

The database also contained 88 AD subjects with CSF

data. The data from the AD subjects were used only for

exploratory purposes to visualize differences between AD

andMCIsubjects.ThedatafromADsubjectswerenotused

inthecurrentmodelbuildingexercise.Adescriptionofthe

objective behind each stage of the modelling procedure

described below is provided in Table S1.

Data analysis software

Data set preparation was performed using SAS® Version

9.1.3 (SAS Institute Inc.,Cary,NC,USA).Data set exploration

and visualization were performed using S Plus® 6.0 profes-

sional release 2 software (Insightful Corporation, Seattle,

WA, USA). ADAS-cog and CSF biomarker data were mod-

elled using extended least squares regression using

NONMEM® VI in combination with the Intel FORTRAN 10

compiler [17].

Selection of the structural model for ADAS-cog

data

The model-building exercise employed log-transformed

data using the first-order conditional estimation method

(FOCE) in NONMEM®.It is known that linear models are not

sufficient for portraying cognitive decline in disease pro-

gression [18,19].The use of logistic curves to describe this

non-linearity in cognitive decline is well accepted [13, 14,

19–22] and these functions offer the advantage that the

modelpredictionsdonotfalloutsidetheboundedscaleof

0to70forADAS-cog.Tojustifythechoiceofthenon-linear

structural model, simpler linear and non-linear models

were also tested (see Results).

A sequence of logistic models [23] was tested and

these models allowed the progression rate to be the

fastest around the inflection point of 42 points on the

ADAS-cog scale [13, 14, 20, 24]. The generalized logistic

model [23] that represents the rate of disease progression

is as follows:

dADAS-cog

dt

A

D DAS-cog

r ADAS-cog

ADAS-cog

ADAS-cog

= × −⎛

⎝

⎞

⎠

⎡

⎣⎢

⎤

⎦⎥

α

β

γ

1

max

;

ADAS-cog ( ) 0

0

=

(1)

where,r is the rate parameter controlling disease progres-

sion,ADAS-cogmaxis fixed at 70,ADAS-cog0is the baseline

score at time zero and a,b and g govern the shape of the

progression curve and also control the inflection point.In

the MCI database there are only four data points (4/1036:

0.4%) with ADAS-cog scores greater than 42 and therefore

estimating any of the shape parameters maybe difficult

withthecurrentdataset(seeResults).Threedifferentlogis-

tic models were tested:(a) in the first model a and b were

fixed at 1, while g was fixed at 0.667, (b) in the second

modelaandgwerefixedat1,whilebwasfixedat2.39and

(c) in the third model b and g were fixed at 1,while a was

fixed at 1.52. Since the relationship between inflection

point and the shape parameters can be derived [23], the

fixed shape parameters in each model allow the inflection

point to be 42, which is in line with the literature derived

value [13, 14, 20, 24] (Table 1).The three models are non-

nested and have the same number of parameters. The

selection of the structural model was therefore guided by

AIC criteria and the model with the lowest AIC value was

considered the base structural model.

Inter-subject variability on baseline ADAS-cog was

evaluated using a log normal distribution because the

parameter had to be constrained to a value greater than

zerowithitsdistributionskewedtotheright.Theapparent

coefficient of variation for inter-individual variability in

baseline ADAS-cog was computed as the square root of

omega (w). Inter-individual variability on the rate param-

eter r was evaluated using an additive-error model.Rate of

progression can be either positive or negative (disease

can worsen or improve over time) in MCI subjects. It is

therefore important to use an additive-error model for

M.N.Samtani et al.

148/

75:1/ Br J Clin Pharmacol

Page 4

parameter r, so that both types of progression can be

captured.The coefficient of variation for inter-individual vari-

ability on the r parameter was computed as 100% ¥ w/

population estimate.Since ADAS-cog scores were log trans-

formed, an additive error model was used to describe the

residual variability. The scores are non-negative and were

increasinglyvariableasthevalueofthescoresincreased.Both

these characteristics are captured adequately using the

log-transform both sides approach for the residual error

[25,26].This approach involves logarithmic transformation

of both the observed data and model predictions, which

induces normality and allows variance stabilization [25,

26]. The magnitude of the residual variability parameter

was expressed as a standard deviation.

Mixture model for ADAS-cog data

It was observed that the inter-individual variability esti-

mates for the progression rate parameter r in the base

structural model was quite high (>100% coefficient of

variation).The high variability is also visible in the longitu-

dinal ADAS-cog scores in MCI subjects (see Results). This

led to the hypothesis that the MCI population consists of a

mixture of two sub-populations and mixing of these non-

homogenous populations led to high inter-individual vari-

ability.Thesetwosub-populationscouldrepresentfastand

slow progressers. Slow progressers were defined as those

having a lower r parameter and lower baseline ADAS-cog

(andviceversaforfastprogressers).Totestthepossibilityof

two sub-populations, mixture modelling, as implemented

in NONMEM® VI [27–29],was applied to the ADAS-cog data.

To allow flexibility, residual variability was allowed to vary

between the two sub-populations.

Mixture models for baseline CSF biomarker

data

Two-component mixture models were also fitted sepa-

rately for each of the baseline CSF biomarker data (CSF

Ab1–42, tau, p-tau181P, and p-tau181P: Ab1–42 ratio) under

the assumption that the data are sampled from 2 different

normal distributions.Since there is a single baseline mea-

surementpersubject,onlyonelevelofrandomeffectswas

implemented using an additive error model. Both tau

markers had right skewed distributions and therefore

CSF tau, p-tau181P, and p-tau181PA : b1–42 ratio were log

transformed before analysis to satisfy the normality

assumption. The thresholds for p-tau181p, Ab1–42 and

p-tau181p: Ab1–42 were based on the densities of their

bimodal distribution.The threshold is taken as the lowest

point in the trough between the two peaks where the

density curves of the two distributions for the mixture

population meet (see Results).

Computation of % correct classification

statistics

It was conjectured that MCI subjects with non-pathologic

CSF could be the slow progressers, while subjects with

pathologicCSFcouldbethefastprogressers.ForCSFAb1–

42,subjects below the critical threshold (identified by the

mixture model above) were considered having pathologic

CSF. In contrast, for CSF p-tau181P and p-tau181P: Ab1–42

ratio,subjectsabovethecriticalthresholdfromtherespec-

tive mixture models were considered to have pathologic

CSF. To assess whether there could be a correlation

between ADAS-cog progression and CSF status,% correct

classification(%CC)[27]statisticswerecomputedbetween

each subject’s post hoc estimate of sub-population assign-

ment from the ADAS-cog mixture model and the CSF

sub-population category based on the CSF biomarker

threshold.The%CCwascomputedforeachCSFbiomarker,

where CC is either pathologic CSF corresponding to fast

progresser status or non-pathologic CSF corresponding to

slow progresser status.

CSF biomarkers as covariates in the ADAS-cog

base structural model

The %CC was high for CSF Ab1–42, p-tau181P, and

p-tau181P: Ab1–42 ratio and therefore an assessment was

made whether these could serve as categorical covariates

in the ADAS-cog base structural model. The optimal

threshold for dichotomizing these biomarkers into cat-

egorical covariates was fixed based on the mixture model

for these biomarkers described earlier. Three separate

ADAS-cog models were fitted, one with CSF Ab1–42, one

with p-tau181P and another with p-tau181P: Ab1–42 ratio,

Table 1

Summary of structural models

Model description Progression rateInflection pointFixed parameter†Number of qs AIC value

Logistic 1

dADAS-cog

dt

r ADAS-cog

ADAS-cog

70

= ×−

⎡

⎣⎢

⎤

⎦⎥

1

γ

70

1+ γ

g = 0.6672

-1126

Logistic 2

dADAS-cog

dt

rADAS-cog

ADAS-cog

70

= ×−()

⎡

⎣⎢

⎤

⎦⎥

1

β

70

1

+

1

β

β

β

⎛

⎝⎜

⎞

⎠⎟

b = 2.392

-1128

Logistic 3

dADAS-cog

dt

rADAS-cog

ADAS-cog

70

= ×−

⎡

⎣⎢

⎤

⎦⎥

α1

α

1

α

×

+

70

a = 1.522

-1129

†In all three models the fixed parameter corresponds to an inflection point at an ADAS-cog score of 42.

Disease progression model in MCI subjects

Br J Clin Pharmacol/

75:1/149

Page 5

and in all models the biomarkers were formulated as cat-

egoricalcovariatestoaffectbothbaselineADAS-cogandr.

These models had the same number of parameters and

selectionofthemoreoptimalbiomarkerasacovariatewas

based on the AIC criteria. For completeness, once the

optimal biomarker was selected, it was also tested as a

continuouscovariateonbaselineADAS-cogandrusing(a)

linear function,(b) power function and (c) log linear func-

tion and the choice of the functional form of the covariate

was also based on AIC.The categorical covariate formalism

also offers two other advantages that were also tested:(a)

the assumption can be tested whether the slow progress-

ers are non-progressers and (b) the assumption can be

tested whether the residual variability between the two

sub-populations is sufficiently differentThe model chosen

after incorporation of CSF biomarkers in the ADAS-cog

model will be referred to as the base reference model.

Assessment of applicability of AD model

covariates to the MCI population

The development of the base reference model led to the

observation that there are only 129 progressers in the

current dataset. A covariate search on such a small data-

base could cause identification of incorrect covariate rela-

tionships due to random noise.Moreover,such an analysis,

which could be associated with low power, may identify

spurious and/or exaggerated covariate relationships [30].

Therefore, further covariate search was guided by prior

knowledgerelatedtothisdiseasearea.Previousanalysisof

covariate relationships in the ADNI population has sug-

gested that baseline ADAS-cog is affected by baseline hip-

pocampal volume, baseline ventricular volume and years

since dementia onset at baseline [13,14].Furthermore,the

r parameter is associated with baseline age, APOE4, base-

line cholesterol and baseline Trail B test [12, 13, 31]. Since

the MCI progressers identified in the current analysis have

AD pathology (high p-tau181P: Ab1–42 ratio),the relevance

ofthesepreviouslyknownADcovariateswasalsotestedin

the MCI population (except years since dementia onset,

which is not relevant to MCI). NONMEM® VI was used to

optimize and finalize the covariate model In the model,

continuous covariates were modelled using a power func-

tion after normalization by the typical reference value

(population median), while categorical covariates were

introduced as fractional shifts [32]. All of the influential

covariates from the previous AD analysis [13, 14] were

added to the base reference model using the appropriate

functional form [32]. Covariates introduced into the full

modelwerethentestedusingbackwardelimination,apro-

ceduredescribedbyWahlbyet al.[33],andtheobjectiveof

this analysis was to develop the most parsimonious cova-

riate disease progression model in MCI.

Finally, to assess the precision and stability of the final

model,theparameterestimatesweresubjectedtointernal

model evaluation [13, 14]. The evaluation consisted of a

non-parametric bootstrap and a visual predictive check

[13,14,34,35].Bootstrap analysis was performed using the

package Perl Speaks NONMEM®,version PsN-3.1.0 [34].

Results

Subject characteristics

The characteristics of the ADNI MCI subjects with CSF

information are shown in Table 2. Petersen et al. have

recentlyreportedthedemographicandbiomarkercharac-

teristics of all the 398 MCI subjects recruited in the ADNI

trial [36].The characteristics of the 198 MCI subjects with

CSFinformationinthecurrentanalysis(Table 2)arealmost

identical to the statistics for the full set of 398 MCI subjects

(similar distribution for age, APOE, gender, educational

statusandcognitivetests).Thisindicatesthatthesubsetof

MCI subjects with CSF information represents a represen-

tative sample of the larger population.This subset of MCI

subjects was between the ages of 55 to 89 years (mean ?

standard deviation [SD] 75 ? 8). The subjects had, on an

average 16 years (?3 SD) of education. Ninety-eight sub-

jects (49.5%) had a family history of dementia with at least

one parent having the disease. There was an apparent

patternformaternaltransmissionofthediseasesince77of

the 98 subjects with a family history had mothers with

dementiawhichisconsistentwithearlierreportsintheAD

literature [13].54% of MCI subjects were APOE e4 carriers,

where 43% had one e4 allele and 11% had two e4

alleles.MCI subjects also had relatively high serum choles-

terol, with the mean cholesterol concentration being

198 mg dl-1(? 43 SD), which is close to the high choles-

terol cut-off of ?200 mg dl-1.

Choice of structural model

The results from the logistic structural model selection

process are shown in Table 1. AIC values for the various

structural models indicate that logistic model 3 with a

shape parameter was the most suitable (i.e. lowest value

among the three models tested).This model form has also

been reported to describe AD disease progression quite

well [13, 14].To understand the behaviour of these struc-

tural models, the progression rate was plotted as a func-

tion of the current ADAS-cog score using the parameter

estimates from each model. The results are presented in

Figure S1, which indicate that the three separate expo-

nents (a,b and g) control both the inflection point and the

initial shape of the curvature characterizing the relation-

ship between progression rate and ADAS-cog.The model

with a shape parameter had greater flexibility at low

ADAS-cog scores,which is particularly relevant to the MCI

population (Figure S1). It is therefore reassuring that this

function was identified here and in previous work [13,14]

as a suitable structural model for describing ADAS-cog

progression.

A linear model for disease progression was also tested

and it resulted in an AIC value of -1094, which signifies

M.N.Samtani et al.

150/

75:1/Br J Clin Pharmacol

Page 6

poorer model fit compared with the logistic models

(Table 1). A simplified logistic function was tested next,

which does not have a shape parameter (i.e.characterized

by an inflection point at half-maximal score of 35) and this

model produced an AIC of -1125.It should be noted that

all the logistic models that had a shape factor gave better

AIC values (Table 1) than the simplified logistic model.This

behaviour agrees with the published literature [13,14,20,

24]thattheinflectionpointforADAS-cogiscloseto42and

not at the mid-point of the ADAS-cog scale. Finally, the

logistic model with the a shape parameter that had the

lowest AIC value was also rerun where a was estimated

insteadofbeingfixed.Themodelransuccessfullyandgave

an estimate of a of 1.48 (inflection point = 41.8); which is

very close to the fixed value of a = 1.52 based on prior

knowledge. However, estimating a led to poorer param-

eter precision and therefore a was kept fixed at 1.52 based

on extensive knowledge [13,14,18–22,24] about the tem-

poral nature of cognitive decline to ensure model stability.

In summary, this exercise of testing various structural

modelsconfirmedtheutilityoftheADstructuralmodelfor

the MCI population, which is not surprising since 42% of

the current MCI population converts to AD during the

courseofthestudy.Thelogisticstructuralmodelwiththea

shapeparameterwasthustakenforwardforassessmentof

mixture populations and covariate analysis and is referred

to as the base model.

Mixture model for ADAS-cog

Results from the base model indicated that the between

subject variability for the progression rate parameter was

113% coefficient of variation.This led to the formulation of

a mixture model for MCI ADAS-cog data. The parameter

estimates of the mixture model are shown in Table S2,

Table 2

Summary statistics for ADNI MCI subjects with CSF data

Variable name (abbreviation), units

Mean (? SD) or n (%)

Subjects with

pathologic CSF*

(n = 129)

All subjects

(n = 198)

Subjects without

pathologic CSF†

(n = 69)

Baseline MRI volumetric measures

Ventricular volume (ml)

Hippocampal volume‡ (mm3)

Baseline chemical biomarkers

Serum cholesterol, mg dl-1

Subjects with high cholesterol, ? 200 mg dl-1

CSF Ab1–42

CSF tau

CSF p-tau181P

Log CSF p-tau181P: Ab1–42 ratio

Demographic and genetic factors

Baseline age (AGE), years

Apolipoprotein E genotype status (APOE4)

0 allele

1 allele

2 alleles

Family history of dementia (FHD)

None

Father

Mother

Both

Gender (SEX)

Male

Female

Years of education (EDU) at baseline

Baseline cognitive tests

ADAS-cog

Mini-mental state exam (MMSE)

Trail making test; part B, s

Longitudinal ADAS-cog scores

Baseline

6 months

1 year

1.5 years

2 years

3 years

44.6 ? 24

3146 ? 528

42.3 ? 21

3045 ? 468

48.8 ? 28

3334 ? 583

198 ? 43

87 (44%)

164 ? 55

103 ? 60

35 ? 18

-1.6 ? 0.7

202 ? 45

60 (47%)

134 ? 30

125 ? 63

44 ? 16

-1.14 ? 0.4

192 ? 39

27 (39%)

218 ? 49

62 ? 22

19 ? 5

-2.46 ? 0.4

75 ? 874 ? 7 75 ? 8

92 (46%)

85 (43%)

21 (11%)

40 (31%)

69 (53%)

20 (16%)

52 (75%)

16 (23%)

1 (1.4%)

100 (51%)

21 (11%)

65 (33%)

12 (6.1%)

60 (47%)

12 (9.3%)

48 (37%)

9 (7.0%)

40 (58%)

9 (13%)

17 (25%)

3 (4.3%)

132 (67%)

66 (33%)

16 ? 3

79 (61%)

50 (39%)

16 ? 3

53 (77%)

16 (23%)

16 ? 3

11.7 ? 5

26.9 ? 2

133 ? 73

Mean ? SD (n)

11.7 ? 5 (n = 198)

12.5 ? 5 (n = 190)

12.6 ? 6 (n = 184)

13.5 ? 7 (n = 169)

14.0 ? 7 (n = 158)

15.2 ? 9 (n = 118)

12.7 ? 5

26.8 ? 2

140 ? 74

Mean ? SD (n)

12.7 ? 5 (n = 129)

13.7 ? 5 (n = 125)

14.2 ? 6 (n = 121)

15.6 ? 7 (n = 111)

16.3 ? 7 (n = 106)

17.8 ? 9 (n = 76)

9.9 ? 4

27.1 ? 2

121 ? 69

Mean ? SD (n)

9.9 ? 4 (n = 69)

10.1 ? 5 (n = 65)

9.6 ? 4 (n = 63)

9.6 ? 5 (n = 58)

9.3 ? 5 (n = 52)

10.5 ? 6 (n = 42)

*Subjects with pathologic CSF at baseline: log CSF p-tau181P: Ab1–42 ratio > -1.86. †Subjects without pathologic CSF at baseline: log CSF p-tau181P: Ab1–42 ratio ? -1.86.

‡Average of left and right hippocampal volume.

Disease progression model in MCI subjects

Br J Clin Pharmacol/

75:1/151

Page 7

which indicate that slow progressers have both a lower

progression rate and a lower baseline score.The progres-

sion rate is even slower after accounting for the lower

ADAS-cog scores observed in the MCI population (the

model uses the logistic structural form). The mixing frac-

tion for progressers was 70%, indicating that 30% of the

subjects could be progressing slowly in the MCI popula-

tion.Tounderstandthebiologicalbasisbehindthishetero-

geneity in the MCI population, CSF biomarkers were

assessedforbimodality.Thebimodalityinprogressionrate

could be associated with dichotomy in the distribution of

CSF biomarkers.Therefore, mixture models for CSF biom-

arkers were assessed next.

Mixture models for CSF biomarkers and %CC

statistics

Out of the four CSF candidate markers, three depicted

possible bimodality (Figure 1). CSF total tau exhibited

unimodality and a right skewed distribution, which was

log-transformed to approximate normality (Figure 1C). A

mixture model could not be successfully fitted to the log-

transformed CSF total tau distribution and it was therefore

notconsideredfurtherasacandidatemarker.Forthethree

remainingmarkersamixturemodelwassuccessfullyfitted

and the results of the analysis are presented in Table S3.A

mixtureoftwonormaldistributionswithnearlyequalstan-

dard deviations is bimodal if their means differ by at least

0.012

0.004

0.008

Density

A

Baseline CSF Ab1-42 (pg ml–1)

in MCI subjects

0.0

0100 200300

0.6

0.2

0.4

Density

B

Baseline CSF p-tau181p on a log

scale in MCI subjects

0.0

2.0 2.5 3.03.54.0 4.5

0.8

0.2

0.4

0.6

Density

C

Baseline CSF total tau on a log

scale in MCI subjects

0.0

3456

0.6

0.2

0.1

0.4

0.3

0.5

Density

D

Baseline CSF p-tau181p : Ab1-42 ratio on a log

scale in MCI subjects

0.0

–3–2 –10

Figure 1

Distribution of CSF biomarkers in MCI subjects. The solid and dashed lines (1A, 1B and 1D) represent the density of the sub-populations based on the

parametersoftheCSFmixturemodelswhilethedottedverticallinesarethecut-offthresholdsseparatingthetwosub-populations.(A,B,D)

pathology;,Without disease pathology

,Withdisease

M.N.Samtani et al.

152/

75:1/ Br J Clin Pharmacol

Page 8

twice the common standard deviation [37].This expecta-

tion of bimodality is met for CSF Ab1–42 (Figure 1A), CSF

p-tau181p (Figure 1B) and CSF p-tau181p: Ab1–42 ratio

(Figure 1D) based on the parameter estimates reported in

Table S3. The ability to fit mixture models with distinct

random effect parameters is dependent upon the nature

of the underlying mixture (i.e. how close are the sub-

population means and how much data are available per

sub-population). Attempts to fit separate random effects

forCSFsub-populationsledtomodelinstabilitywhichwas

reflected in higher imprecision for model parameters.Thus

the two sub-populations for the CSF biomarkers were

assumed to have the same variances (Table S3) as is com-

monly done in the implementation of mixture models in

NONMEM® [28, 29]. Furthermore, fitting the model without

subpopulations to the baseline CSF dataset for p-tau181p,

Ab1–42 and p-tau181p: Ab1–42 ratio resulted in much

worse fit (based on AIC and likelihood ratio test).

The mixing proportion for p-tau181P, p-tau181P: Ab1–42

and Ab1–42 were 55%, 64% and 75% respectively and

these are close to the 70% mixing proportion for the

ADAS-cog mixture model.The thresholds determined for

p-tau181P,p-tau181P: Ab1–42 and Ab1–42 based on the den-

sities of these bimodal distribution were 29 pg ml-1(log

scale 3.37),0.156 (log ratio -1.86) and 198 pg ml-1respec-

tively and these thresholds are indicated in Figure 1.Based

on these threshold values, the population was dichoto-

mized and the %CC statistic was computed for each

marker using the post hoc estimate of sub-population

assignment from the ADAS-cog mixture model.The %CC

for p-tau181P, p-tau181P: Ab1–42, Ab1–42 were 68%, 73%

and 71% respectively. Since the %CC for p-tau181P,

p-tau181P: Ab1–42andAb1–42wererelativelyhigh(~70%),

all three markers were pursued further as potential covari-

ates in the ADAS-cog base model.Since p-tau181P: Ab1–42

ratio gave the highest %CC statistic,the contingency table

between CSF status and progresser status from the

mixture model is reported in Table S4.

CSF biomarkers as covariates for ADAS-cog

disease progression

CSFAb1–42wasincorporatedasacategoricalcovariateon

both baseline ADAS-cog and r parameter,which produced

an AIC value of -1163. The AIC value with the model

parameterizing CSF p-tau181P as the covariate was -1167.

Finally, theAICvalue for

p-tau181P: Ab1–42 ratio as a covariate on the same param-

eters was -1181.This suggested that the ratio of the two

biomarkers may carry more information than a single

biomarker alone and it was chosen as the CSF-related

covariate in the ADAS-cog model. For completeness, the

ratio of p-tau181P: Ab1–42 was also tested as a continuous

covariatethroughalinear,log-linearorpowerrelationship,

which yielded AIC values of -1165, -1179 and -1173.CSF

p-tau181P: Ab1–42 ratio thus produces the lowest AIC value

when it is formulated as a categorical covariate.This sug-

themodelwithCSF

geststhattheseCSFendpoints(p-tau181PandAb1–42)may

serve as a threshold between occult and measureable

disease progression. It is noteworthy that simply adding

these two parameters to the ADAS-cog base model,

p-tau181PA : b1–42 affecting baseline ADAS-cog and r,

reduced the minimum value of the objective function by

56 points which is highly significant (P < 0.00001). It was

also noticed that the estimate of the r parameter in the

slow progressers (log p-tau181P: Ab1–42 ? -1.86) was

0.005,whichisveryclosetozero.Therefore,theassumption

was tested whether these subjects represent non-

progressers with a typical r parameter value of zero. This

simplification led to an increase in the minimum value of

the objective function by 0.4 points.Therefore, the model

reduction by one parameter did not lead to a significant

changeinthefit.Thus,basedonthisanalysis,fastprogress-

ers will be referred to as progressers,while slow progress-

ers will be referred to as non-progressers. Progressers are

defined as subjects with log p-tau181P: Ab1–42 > -1.86,

while non-progressers are defined as subjects with log

p-tau181P: Ab1–42 ? -1.86.

Thenon-progressers,becauseoftheirsmallsamplesize

(n = 69),were also constrained to have their etas (h:devia-

tionofanindividualparameterfromthepopulationmean)

sampled from the same w distribution as that of the pro-

gressers. However, to allow greater flexibility the residual

variability was allowed to vary between progressers and

non-progressers. Addition of one extra residual error

parameter led to an improvement of 18 objective function

points.It was also noticed that the SD of the residual error

for the non-progressers (0.30) was somewhat larger than

that for the progressers (0.24). This is because the ADAS-

cog score for the non-progressers fluctuates more widely

around a relatively steady value.This model in which the

non-progressers had a typical progression rate parameter

ofzero,possessedalowerbaselinescoreandwereallowed

to have a different residual variability was considered the

base reference model and was tested further for covariate

model building.

Final covariate model and model verification

Further covariate model building proceeded via a full

model/backward elimination procedure in NONMEM® VI.

The procedure identified only two new covariates in the

model, which were hippocampal volume and the Trail B

test (Table 3).Plots of baseline ADAS-cog h for progressers

andnon-progressersvs.hippocampalvolumeshowedthat

the baseline score was dependent on this volumetric

marker for both these populations.Furthermore,since the

non-progressers represent a smaller fraction of the whole

population (n = 69) only a single hippocampal volume

related parameter was fitted for baseline ADAS-cog in the

entire MCI population.The h for the r parameter vs.Trail B

test score in non-progressers did not show any trend and

therefore this covariate influences only the progressers.

Disease progression model in MCI subjects

Br J Clin Pharmacol/

75:1/153

Page 9

Figure S2 shows the goodness of fit plots for the final

model and Table 3 provides the estimates from the final

population based disease progression model. The results

ofthenon-parametricbootstrapanalysis(Table 3)support

the parameter estimates of the final model.The parameter

estimates are similar to the median value obtained from

thebootstraptechniqueandarecontainedwithinthe90%

confidence interval.The observed scores,the visual predic-

tive check, and median model prediction vs. time are dis-

played in Figure 2.These results confirm that the model is

able to describe the ADAS-cog temporal profiles in MCI

subjects since the majority of the observations fall within

the 90% prediction intervals (Figure 2).

Discussion

Characteristics of MCI population identified

based on model based analysis

ThreekeycharacteristicsoftheMCIpopulationemerge:(a)

theMCIpopulationpotentiallyrepresentsamixtureoftwo

sub-populations, (b) among the MCI progressers, some of

the subjects progress at a relatively slower rate likely due

to additional factors such as preserved executive function

and(c)amongthenon-progressers32subjects(16%ofthe

MCI population) had a value of the r parameter that was

negative,which indicates that some non-progressers may

havetheabilitytorevertbacktonormalstatus.Thistypeof

variability in the clinical course of MCI subjects has been

described previously [2].

Rationale for testing CSF biomarkers as

covariates

In the previous AD analysis [13,14],the CSF data were not

used since they were available in only 88 subjects.

However, 198 MCI subjects had CSF information, which

represents a reasonable size sample for investigating CSF

biomarkers as covariates. At the current time, diagnosis

of AD requires presence of dementia. However, there has

been speculation that individuals who are bound to

develop AD can be identified earlier using CSF biomarkers

[38].Abandp-tau181Pareanintegralpartofdiseasepathol-

ogy and it is interesting that progression on a clinical scale

(ADAS-cog)ismirrored in

p-tau181P: Ab1–42.The critical threshold identified for this

ratio in the current analysis is -1.86 (untransformed scale

0.156). MCI subjects below this critical threshold do not

appear to exhibit disease progression (Figure 2A). This

probably indicates that these subjects either do not have

the disease pathology or the pathologic cascade has not

started yet.

theratiooflog CSF

Role of APOE and cholesterol

In previous models of AD progression both APOE e4 and

serum cholesterol have been identified as covariates that

predicted faster disease progression [13,14,31].Ab,APOE

and cholesterol are linked with one another [39–41] since

the APOE e4 allele is linked with disturbances in Ab and

cholesterol metabolism. In the current analysis, APOE e4

and serum cholesterol were not identified as statistically

significant covariates. Instead, Ab1–42 and p-tau181P are

covariates in the model. If the entire MCI population is

Table 3

Population parameters and the precision of the parameters using nonparametric bootstrap

Parameter*

Original dataset

Estimate

Bootstrap replicates (n = 1000)

Median90% CI90% CI

qADAS-cog0

qHVOL

qCSF

qr

qTRAB

Inter-subject variability (% coefficient of variation)†

ADAS-cog0

r

Residual variability (SD)

Population with pathologic CSF‡

Population without Pathologic CSF‡

11.3

-0.863

0.827

0.042

0.621

10.7

-1.10

0.747

0.034

0.379

11.9

-0.629

0.907

0.049

0.863

11.3

-0.878

0.827

0.041

0.634

10.7

-1.11

0.756

0.034

0.394

11.9

-0.627

0.908

0.048

0.862

32.2

69.4

28.5

48.0

35.9

90.8

31.8

69.4

28.0

48.3

35.4

92.4

0.237

0.300

0.207

0.269

0.267

0.331

0.233

0.298

0.208

0.269

0.269

0.332

*These equations describe the relationships between covariates and the typical value (TV) of the parameters in the final model:

×(

TV rCSF

r FLAG

=×θ

109

TV ADAS-cog

HVOL

3115

ADAS-cog

)

CSF

csf

HVOL

θ

00

=×

)

θθ

TRAB

TRAB

×(

θ

where; csf is a 0/1 exponent and CSFFLAGis a 1/0 flag variable depending on sub-population with/without pathologic CSF respectively. HVOL, CSF and TRAB refer to hippocampal

volume, cerebrospinal fluid and Trail B test respectively. †Between the base model and final covariate model the inter-subject variability SD estimates improved from 39.5% and 113%

to 32.2% and 69.4% coefficient of variation respectively. ‡Population with pathologic CSF corresponds to log CSF p-tau181P: Ab1–42 ratio > -1.86; population without pathologic

CSF corresponds to log CSF p-tau181P: Ab1–42 ratio ? -1.86.

M.N.Samtani et al.

154/

75:1/ Br J Clin Pharmacol

Page 10

stratified by either APOE or cholesterol status (Figures 3A

and 4A respectively) there is an evident trend that these

factors affect progression rate. However, if the population

is first dichotomized by p-tau181P: Ab1–42 CSF status and

the influence of APOE and cholesterol are assessed, then

the trend disappears (Figures 3B, 3C and 4B, 4C respec-

tively).Furthermore,84% (89/106) of APOE e4 carriers have

the pathologic CSF ratio.Similarly 69% (60/87) of MCI sub-

jects with high cholesterol have pathologic CSF ratio.Thus

APOE e4,high cholesterol and high p-tau181P: Ab1–42 ratio

are likely correlated with one another. This probably also

explains why high cholesterol and APOE e4 were signifi-

50

40

30

A

Time (years)

ADAS-cog score

20

10

0

01234

50

40

30

B

Time (years)

ADAS-cog score

20

10

0

01234

Figure 2

Results of the stratified visual predictive check; x-axis are jittered for clarity.Open symbols are observed data while lines and shaded areas represent the

median and 90% prediction intervals. (A) Non-progressers without pathologic CSF [log CSF p-tau181P: Ab1–42 ratio ? -1.86] and (B) progressers with

pathologic CSF [log CSF p-tau181P: A b1–42 ratio > -1.86]

Time (years) Time (years)

18

16

14

12

10

B

Mean ADAS-cog score

8

0.0 0.5 1.0 1.5 2.0

18

16

14

12

10

C

Mean ADAS-cog score

8

0.0 0.51.0 1.52.0

18

16

14

12

10

A

Time (years)

Mean ADAS-cog score

8

0.00.51.0 1.5

2.0

Figure 3

Influence of APOE may no longer be apparent once the data are dichotomized by CSF status.(A) entire MCI population,(B) progressers with pathologic CSF

[log CSF p-tau181P: Ab1–42 ratio > -1.86] and (C) non-progressers without pathologic CSF [log CSF p-tau181P: Ab1–42 ratio ? -1.86].APOE e4 was dichoto-

mized into carrier (one or two alleles) and non-carrier status.Error bars represent standard error (SE) and lines are simple linear regression through the data

to allow visualization of trends.? APOE4 non-carrier;? APOE4 carrier

Disease progression model in MCI subjects

Br J Clin Pharmacol/

75:1/ 155

Page 11

cant in the AD analysis where CSF biomarkers were not

tested [13, 14, 31]. The p-tau181P: Ab1–42 ratio was highly

significant in the MCI disease progression model and

maybe a more useful covariate than APOE e4 and

cholesterol.

Other comparisons between MCI and AD

progression models

Hippocampal volume and theTrail B test have been previ-

ously identified as influential covariates in AD [13, 14],

which are equally significant in the current MCI analysis.To

allow visualization of important covariate effects, some

simple diagnostics were created (Figures 5, 6, and Figure

S3).For these plots the important covariates were dichoto-

mized (> Median and ? Median) to create roughly equal

groups and the mean ADAS-cog was plotted as a function

of this newly created categorical variable. Hippocampal

volume was associated with baseline scores for both pro-

gressers and non-progressers, (Figure 5A and 5B). This

finding for hippocampal volume is consistent with the lit-

erature where cognitive decline was associated with hip-

pocampal atrophy [42]. Additionally, a longer completion

19

16

13

10

A

Time (years)Time (years) Time (years)

Mean ADAS-cog score

7

0.0 0.51.0 1.52.0

19

16

13

10

7

B

Mean ADAS-cog score

0.00.5 1.0 1.52.0

19

16

13

10

7

C

Mean ADAS-cog score

0.00.5 1.0 1.5 2.0

Figure 4

Influence of cholesterol is no longer apparent once the data are dichotomized by CSF status.(A) entire MCI population,(B) Progressers with pathologic CSF

[logCSFp-tau181P: Ab1–42ratio>-1.86]and(C)non-progresserswithoutpathologicCSF[logCSFp-tau181P: A-1b1–42ratio?-1.86].Totalserumcholesterol

was dichotomized into high cholesterol (?200 mg dl-1) and normal cholesterol (<200 mg dl-1).Error bars represent standard error (SE) and lines are simple

linear regression through the data to allow visualization of trends.(A,B,C) ? normal cholesterol;? high cholesterol

25

22

19

16

13

A

Time (years) Time (years)Time (years)

Mean ADAS-cog score ± SE

10

7

0.0 0.51.01.5

2.0

25

22

19

16

13

B

Mean ADAS-cog score ± SE

10

7

0.0 0.51.0 1.52.0

25

22

19

16

13

C

Mean ADAS-cog score ± SE

10

7

0.0 0.5 1.0 1.52.0

Figure 5

(A) Influence of hippocampal volume on baseline disease score for non-progressers,(B) impact of hippocampal volume on baseline score for progressers

and (C) disease progression rate affected by Trail B test in progressers. Covariates were dichotomized to create roughly equal groups (> Median and ?

Median) in each panel. Median hippocampal volume in progressers and non-progressers were 3045 and 3334 mm3respectively. Median Trail B test in

progressers was 109 s.Error bars represent standard error (SE) and lines are simple linear regression through the data to allow visualization of trends.(A,B)

? low hippocampal volume;? high hippocampal volume.(C) ? lower Trail B test time;? higher Trail B test time

M.N.Samtani et al.

156/

75:1/ Br J Clin Pharmacol

Page 12

time on theTrail B test was associated with faster progres-

sion for subjects with pathologic CSF (Figure 5C), which

indicates that patients with poor executive function

progress rapidly.

Two additional covariates (ventricular volume and age)

that were identified previously in AD [13, 14] were not

statistically significant in the MCI analysis.The inability to

identify ventricular volume in the current analysis is prob-

ably related to the narrow range of the baseline data in

MCI. There is an apparent trend for the influence of ven-

tricular volume on baseline ADAS-cog (Figure S3) but this

trend does not reach statistical significance. As far as the

influence of age is concerned,it does appear that there is a

differential effect of this covariate on AD vs. MCI subjects

(Figure 6).It seems that if the onset of AD dementia occurs

at an early age then the form of the disease is rather

aggressive and progression is quite rapid (Figure 6A).

However,in the MCI population,the onset of dementia has

not yet occurred and age does not appear to influence

disease progression substantially (Figure 6b).

Summary of findings: application of CSF

biomarkers for trial enrichment

The CSF findings from this analysis match with the ADNI

information about the number of MCI subjects who have

either converted (n = 84) or not converted (n = 114) to AD.

The information about converters and non-converters

from ADNI, as a function of CSF biomarker status, is

depictedinFigure 7.TheresultsindicatethattheCSFinfor-

mation, at the individual level, has good negative predic-

tive value i.e. 58 out of the 69 (84%) subjects with log

p-tau181P: Ab1–42 ratio ? -1.86 have still not converted to

AD.Moreover,itisalsoreassuringtoseethat87%(73/84)of

the converters have high log p-tau181P: Ab1–42 ratio

(> -1.86).In contrast,56 out of the 129 subjects (43%) with

log p-tau181P: Ab1–42 ratio > -1.86 have still not converted

to AD. These subjects likely will either (a) eventually

develop AD as the 2–3 follow-up period in the current

database may not be long enough or (b) it is also possible

that these subjects have other protective factors (e.g.pre-

served executive function) that temporarily slow down

their progression rate.Thus these CSF biomarkers may not

precisely predict clinical conversion to AD. They can,

however, be quite useful in excluding those subjects who

have a low likelihood of exhibiting disease progression

within a 2–3 year time frame of a clinical trial. Since non-

progressing subjects may cause noise in an MCI clinical

trial (higher residual error), it may be prudent to exclude

them. The utility of CSF biomarkers as a trial enrichment

tool has recently received regulatory attention in a qualifi-

cation opinion issued by the European Medicines Agency

[43]. Furthermore, there is at least one pharmaceutical

company that is using this technique for population

enrichment [44] and there are two distinct reasons for

excluding these patients in a prodromal AD study:(i) they

arelikelytoremainstableonbothplaceboandactivearms

and(ii)thesesubjectslikelydonothaveAbandtauabnor-

malities and may not benefit from a therapy directed

towards plaque and tangle pathology.

Two recent publications report results that are quite

compatible with the current analysis.First,Buchhave et al.

reportaclinicalstudyfromSwedenwithmedianfollow-up

time of 9.2 years in 137 MCI patients [45]. In this study

baseline p-tau181p: Ab1–42 ratio again exhibited bimodal-

ity and 90% of the patients with pathologic CSF biomarker

levels (high p-tau181p and low Ab1–42) developed AD in

9–10 years. Secondly, Snider et al. report another smaller

study with 49 MCI subjects with longitudinal profile for

clinical dementia rating-sum of boxes which is another

cognitive and functional end point [46].Their results also

34

28

25

22

19

16

13

10

0.00.5

0.5

1.01.5 2.0

31

34

28

25

22

19

16

13

10

0.01.01.5 2.0

31

A

Time (years)

Mean ADAS-cog score

B

Time (years)

Mean ADAS-cog score

Figure 6

Differential effect of age on (A) AD subjects vs.(B) MCI progressers.For dichotomization the median age for AD subjects and MCI progressers were 76 and

74 years respectively.Error bars represent standard error (SE) and lines are simple linear regression through the data to allow visualization of trends.(A,B)

? lower age group;? higher age group

Disease progression model in MCI subjects

Br J Clin Pharmacol/

75:1/ 157

Page 13

show that high p-tau181p and low Ab1–42 quantitatively

predicts rapid progression for cognitive decline.Thus the

current analysis and previous results qualify the utility of

CSFbiomarkersforbeingpredictiveoftherateofcognitive

decline on continuous scales rather than just the dichoto-

mous outcome of conversion to AD.

It should be pointed out that MCI clinical trials have

historically used categorical or time to event outcomes as

their primary analysis [47]. Recently Donohue et al. have

suggested that continuous assessment of disease severity

may be more efficient because mixed-effects models use

all available data, which make them more robust [47].

Donohue et al.have also shown that trials with continuous

outcomes have greater power on average than those with

a dichotomous outcome [47]. Thus, the mixed effects

disease progression model presented in the current analy-

sis could also find utility in analyzing data from pivotal

efficacy trials in MCI.

In summary, this work provides an integrated model-

based analysis of disease progression in MCI subjects.This

modelallowsidentificationofsub-populationssuitablefor

trial enrichment and could represent a useful tool for effi-

cient trial design through clinical trial simulations. In par-

ticular, CSF biomarkers can be useful for excluding those

MCI subjects who have a low likelihood of exhibiting

disease progression on both continuous and categorical

end points. Furthermore, continuous end points may be

more suitable than categorical endpoints since they have

the potential to increase the statistical power of clinical

trials.

One of the obstacles for implementing the trial enrich-

ment approach is the variation in biomarker measure-

ments observed between studies and laboratories. Even

though these biomarker distributions show bimodality at

baselineinMCIstudies[5,45]theabsolutevaluesforthese

biomarkers can be quite different. This variation is prob-

ably the result of differences in CSF sample handling tech-

niques, analytical procedures and analytical kits/reagents.

Standardizationoftheseproceduresmayreducethevaria-

tion and increase the utility of these CSF biomarkers.Cur-

rently, there are at least three quality control and

standardizationinitiatives[48–50]underwaythatwilllikely

help with harmonization of CSF biomarker measurements.

Competing interests

The authors of this manuscript are employees of Johnson

& Johnson Pharmaceutical Research & Development

(JnJPRD) and own JnJ stock.

The authors are sincerely grateful to all members of the

Advanced Modeling & Simulation Department at JnJPRD for

100

80

60

A

Baseline CSF Ab1-42 in MCI

subjects (pg ml–1)

Baseline CSF p-tau181p in MCI

subjects (pg ml–1)

40

20

50100 150200250 300

100

80

60

B

Baseline CSF Ab1-42 in MCI

subjects (pg ml–1)

Baseline CSF p-tau181p in MCI

subjects (pg ml–1)

40

20

50 100150200 250 300

Figure 7

Relationship between CSF p-tau181Pand Ab1–42 for MCI subjects that have (A) not converted to AD and (B) converted to AD. In both panels triangles

representsubjectswithlogCSFp-tau181P: Ab1–42ratio>-1.86,whilesquaresrepresentsubjectswithlogCSFp-tau181P: Ab1–42ratio?-1.86.Inbothpanels

open symbols refer to correct assignment i.e.low ratio subjects who do not convert and high ratio subjects who convert to AD.In contrast,filled symbols

represent incorrect assignment i.e. high ratio subjects who have not converted and low ratio subjects who have converted to AD. (A) ? log CSF

p-tau181p: Ab1-42 ratio > -1.86; ? log CSF p-tau181p: Ab1-42 ratio ?-1.86. (B) ? log CSF p-tau181p: Ab1-42 ratio > -1.86;

?-1.86

log CSF p-tau181p: Ab1-42 ratio

M.N.Samtani et al.

158/

75:1/ Br J Clin Pharmacol

Page 14

theirinsightfulcommentsduringtheconductofthisanalysis.

We are also thankful to Harry Chen (JnJPRD) for formatting

NONMEM® ready files for this analysis.Anna Mendlin (JnJPRD)

provided editorial support for this manuscript.

Datacollectionandsharingforthisprojectwasfundedby

ADNI (National Institutes of Health Grant U01 AG024904).

ADNI is funded by the National Institute on Aging, the

NationalInstituteofBiomedicalImagingandBioengineering,

and through generous contributions from the following:

Abbott, AstraZeneca AB, Bayer Schering Pharma AG, Bristol-

Myers Squibb,Eisai Global Clinical Development,Elan Corpo-

ration, Genentech, GEHealthcare,

Innogenetics, Johnson and Johnson, Eli Lilly and Co.,

Medpace, Inc., Merck and Co., Inc., Novartis AG, Pfizer Inc, F.

Hoffman-La Roche, Schering-Plough, Synarc, Inc., as well as

non-profit partners the Alzheimer’s Association and Alzhe-

imer’s Drug Discovery Foundation, with participation from

the U.S.Food and Drug Administration.Private sector contri-

butions to ADNI are facilitated by the Foundation for the

National Institutes of Health (http://www.fnih.org). The

grantee organization is the Northern California Institute for

Research and Education and the study is coordinated by the

Alzheimer’s Disease Cooperative Study at the University of

California, San Diego. ADNI data are disseminated by the

Laboratory for Neuro Imaging at the University of California,

Los Angeles.This research was also supported by NIH grants

P30 AG010129,K01 AG030514 and the Dana Foundation.

GlaxoSmithKline,

REFERENCES

1 Petersen RC.Mild cognitive impairment clinical trials.Nat

Rev Drug Discov 2003;2:646–53.

2 Petersen RC,Roberts RO,Knopman DS,Boeve BF,Geda YE,

Ivnik RJ,Smith GE,Jack CR Jr.Mild cognitive impairment:ten

years later.Arch Neurol 2009;66:1447–55.

3 Gauthier S,Reisberg B,Zaudig M,Petersen RC,Ritchie K,

Broich K,Belleville S,Brodaty H,Bennett D,Chertkow H,

Cummings JL,de Leon M,Feldman H,Ganguli M,Hampel H,

Scheltens P,Tierney MC,Whitehouse P,Winblad B.

International Psychogeriatric Association Expert Conference

on mild cognitive impairment.Mild cognitive impairment.

Lancet 2006;367:1262–70.

4 Brooks LG,Loewenstein DA.Assessing the progression of

mild cognitive impairment to Alzheimer’s disease:current

trends and future directions.Alzheimers Res Ther 2010;2:

1–9.

5 De Meyer G,Shapiro F,Vanderstichele H,Vanmechelen E,

Engelborghs S,De Deyn PP,Coart E,Hansson O,Minthon L,

Zetterberg H,Blennow K,Shaw L,Trojanowski JQ,

Alzheimer’s Disease Neuroimaging Initiative.

Diagnosis-independent Alzheimer disease biomarker

signature in cognitively normal elderly people.Arch Neurol

2010;67:949–56.

6 Hampel H,Frank R,Broich K,Teipel SJ,Katz RG,Hardy J,

Herholz K,Bokde AL,Jessen F,Hoessler YC,Sanhai WR,

Zetterberg H,Woodcock J,Blennow K.Biomarkers for

Alzheimer’s disease:academic,industry and regulatory

perspectives.Nat Rev Drug Discov 2010;9:560–74.

7 Albert MS,DeKosky ST,Dickson D,Dubois B,Feldman HH,

Fox NC,Gamst A,Holtzman DM,Jagust WJ,Petersen RC,

Snyder PJ,Carrillo MC,Thies B,Phelps CH.The diagnosis of

mild cognitive impairment due to Alzheimer’s disease:

recommendations from the National Institute on

Aging-Alzheimer’s Association workgroups on diagnostic

guidelines for Alzheimer’s disease.Alzheimers Dement 2011;

7:270–9.

8 Brys M,Pirraglia E,Rich K,Rolstad S,Mosconi L,Switalski R,

Glodzik-Sobanska L,De Santi S,Zinkowski R,Mehta P,

Pratico D,Saint Louis LA,Wallin A,Blennow K,de Leon MJ.

Prediction and longitudinal study of CSF biomarkers in mild

cognitive impairment.Neurobiol Aging 2009;30:682–90.

9 Fagan AM,Roe CM,Xiong C,Mintun MA,Morris JC,

Holtzman DM.Cerebrospinal fluid tau/beta-amyloid(42) ratio

as a prediction of cognitive decline in nondemented older

adults.Arch Neurol 2007;64:343–9.

10 Hansson O,Zetterberg H,Buchhave P,Londos E,Blennow K,

Minthon L.Association between CSF biomarkers and

incipient Alzheimer’s disease in patients with mild cognitive

impairment:a follow-up study.Lancet Neurol 2006;5:

228–34.

11 Mattsson N,Zetterberg H,Hansson O,Andreasen N,

Parnetti L,Jonsson M,Herukka SK,van der Flier WM,

Blankenstein MA,Ewers M,Rich K,Kaiser E,Verbeek M,

Tsolaki M,Mulugeta E,Rosén E,Aarsland D,Visser PJ,

Schröder J,Marcusson J,de Leon M,Hampel H,Scheltens P,

Pirttilä T,Wallin A,Jönhagen ME,Minthon L,Winblad B,

Blennow K.CSF biomarkers and incipient Alzheimer disease

in patients with mild cognitive impairment.JAMA 2009;302:

385–93.

12 Visser PJ,Verhey F,Knol DL,Scheltens P,Wahlund LO,

Freund-Levi Y,Tsolaki M,Minthon L,Wallin AK,Hampel H,

Bürger K,Pirttila T,Soininen H,Rikkert MO,Verbeek MM,

Spiru L,Blennow K.Prevalence and prognostic value of CSF

markers of Alzheimer’s disease pathology in patients with

subjective cognitive impairment or mild cognitive

impairment in the DESCRIPA study:a prospective cohort

study.Lancet Neurol 2009;8:619–27.

13 Samtani MN,Farnum M,Lobanov V,Yang E,Raghavan N,

DiBernardo A,Narayan V;Alzheimer’s Disease Neuroimaging

Initiative.An improved model for disease progression in

subjects from Alzheimer’s disease neuroimaging initiative.

J Clin Pharmacol 2012;52:629–44.

14 Samtani MN,Farnum M,Lobanov V,Yang E,Raghavan N,

DiBernardo A,Narayan V.An improved model for disease

progression in subjects from Alzheimer’s disease

neuroimaging initiative [Internet].In:American Conference

on Pharmacometrics (ACoP).San Diego:2011.Available at:

http://www.go-acop.org/sites/default/files/webform/posters/

ACOP-Poster.ppt (last accessed 8 June 2011).

15 Shaw LM,Korecka M,Clark CM,Lee VM,Trojanowski JQ.

Biomarkers of neurodegeneration for diagnosis and

monitoring therapeutics.Nat Rev Drug Discov 2007;6:

295–303.

Disease progression model in MCI subjects

Br J Clin Pharmacol/

75:1/ 159

Page 15

16 Toledo JB,Vanderstichele H,Figurski M,Aisen PS,

Petersen RC,Weiner MW,Jack CR Jr,Jagust W,Decarli C,

Toga AW,Toledo E,Xie SX,Lee VM,Trojanowski JQ,Shaw LM,

Alzheimer’s Disease Neuroimaging Initiative.Factors

affecting Ab plasma levels and their utility as biomarkers in

ADNI.Acta Neuropathol 2011;122:401–13.

17 Boeckman A,Sheiner A,Beal S.NONMEM VI.GloboMax,ICON

Development Solutions:Ellicott City,MD,2007.

18 Mendiondo MS,Ashford JW,Kryscio RJ,Schmitt FA.

Modelling mini mental state examination changes in

Alzheimer’s disease.Stat Med 2000;19:1607–16.

19 Stern Y,Liu X,Albert M,Brandt J,Jacobs DM,

Del Castillo-Castaneda C,Marder K,Bell K,Sano M,Bylsma F,

Lafleche G,Tsai WY.Application of a growth curve approach

to modeling the progression of Alzheimer’s disease.J

Gerontol A Biol Sci Med Sci 1996;51:M179–84.

20 Ashford JW,Schmitt FA.Modeling the time-course of

Alzheimer dementia.Curr Psychiatry Rep 2001;3:20–8.

21 van Belle G,Uhlmann RF,Hughes JP,Larson EB.Reliability of

estimates of changes in mental status test performance in

senile dementia of the Alzheimer type.J Clin Epidemiol

1990;43:589–95.

22 Liu X,Tsai WY,Stern Y.A functional decline model for

prevalent cohort data.Stat Med 1996;15:1023–32.

23 Tsoularis A,Wallace J.Analysis of logistic growth models.

Math Biosci 2002;179:21–55.

24 Stern RG,Mohs RC,Davidson M,Schmeidler J,Silverman J,

Kramer-Ginsberg E,Searcey T,Bierer L,Davis KL.A

longitudinal study of Alzheimer’s disease:measurement,

rate,and predictors of cognitive deterioration.Am J

Psychiatry 1994;151:390–6.

25 Carroll RJ,Ruppert D.Transformations and Weighting in

Regression.New York:Chapman & Hall,1988;115–60.

26 Bonate P.Pharmacokinetic-Pharmacodynamic Modeling and

Simulation.New York:Springer,2006;141–44.

27 Kaila N,Straka RJ,Brundage RC.Mixture models and

subpopulation classification:a pharmacokinetic simulation

study and application to metoprolol CYP2D6 phenotype.J

Pharmacokinet Pharmacodyn 2007;34:141–56.

28 Beal SL,Boeckman AJ,Sheiner LB,eds. NONMEM Users

Guide – Part VI.PREDPP Guide.San Francisco,CA:NONMEM

Project Group,University of California,1992;35–6.

29 Ette EI,Williams PJ,eds. Pharmacometrics:The Science of

Quantitative Pharmacology.New York:Wiley,John & Sons,

Incorporated,2007;723–57.

30 Ribbing J,Jonsson EN.Power,selection bias and predictive

performance of the Population Pharmacokinetic Covariate

Model.J Pharmacokinet Pharmacodyn 2004;31:109–34.

31 Ito K,Corrigan B,Zhao Q,French J,Miller R,Soares H,Katz E,

Nicholas T,Billing B,Anziano R,Fullerton T,Alzheimer’s

Disease Neuroimaging Initiative.Disease progression model

for cognitive deterioration from Alzheimer’s Disease

Neuroimaging Initiative database.Alzheimers Dement 2011;

7:151–60.

32 Ravva P,Gastonguay MR,Tensfeldt TG,Faessel HM.

Population pharmacokinetic analysis of varenicline in adult

smokers.Br J Clin Pharmacol 2009;68:669–81.

33 Wählby U,Jonsson EN,Karlsson MO.Comparison of stepwise

covariate model building strategies in population

pharmacokinetic-pharmacodynamic analysis.AAPS

PharmSci 2002;4:E27.

34 Lindbom L,Philgren P,Jonsson N.PsN-Toolkit-a collection of

computer intensive statistical methods for nonlinear mixed

effect modelling using NONMEM.Comput Methods

Programs Biomed 2005;79:241–57.

35 Holford N.The visual predictive check – superiority to

standard diagnostic (Rorschach) plots.PAGE 2005;14:738

(Abstr.).

36 Petersen RC,Aisen PS,Beckett LA,Donohue MC,Gamst AC,

Harvey DJ,Jack CR Jr,Jagust WJ,Shaw LM,Toga AW,

Trojanowski JQ,Weiner MW.Alzheimer’s Disease

Neuroimaging Initiative (ADNI):clinical characterization.

Neurology 2010;74:201–9.

37 Schilling MF,Watkins AE,Watkins W.Is human height

bimodal? Am Stat 2002;56:223–9.

38 Jack CR Jr,Knopman DS,Jagust WJ,Shaw LM,Aisen PS,

Weiner MW,Petersen RC,Trojanowski JQ.Hypothetical

model of dynamic biomarkers of the Alzheimer’s

pathological cascade.Lancet Neurol 2010;9:119–28.

39 Chauhan NB.Membrane dynamics,cholesterol homeostasis,

and Alzheimer’s disease.J Lipid Res 2003;44:2019–29.

40 Leduc V,Jasmin-Bélanger S,Poirier J.APOE and cholesterol

homeostasis in Alzheimer’s disease.Trends Mol Med 2010;

16:469–77.

41 Shobab LA,Hsiung GY,Feldman HH.Cholesterol in

Alzheimer’s disease.Lancet Neurol 2005;4:841–52.

42 van de Pol LA,Hensel A,Barkhof F,Gertz HJ,Scheltens P,

van der Flier WM.Hippocampal atrophy in Alzheimer

disease:age matters.Neurology 2006;66:236–8.

43 European Medicines Agency.Qualification Opinion of

Alzheimer’s Disease Novel Methodologies/Biomarkers for

BMS-708163.London:EMA,2011;Doc Reference number

EMA/CHMP/SAWP/102001/2011.

44 ClinicalTrial.gov. A multicenter,double blind,

placebo-controlled,safety and tolerability study of

BMS-708163 in patients with prodromal Alzheimer’s disease.

Available at:http://clinicaltrials.gov/ct2/show/NCT00890890

(last accessed 21 May 2012).

45 Buchhave P,Minthon L,Zetterberg H,Wallin AK,Blennow K,

Hansson O.Cerebrospinal fluid levels of b-Amyloid 1-42,but

not of tau,are fully changed already 5 to 10 years before the

onset of Alzheimer dementia.Arch Gen Psychiatry 2012;69:

98–106.

46 Snider BJ,Fagan AM,Roe C,Shah AR,Grant EA,Xiong C,

Morris JC,Holtzman DM.Cerebrospinal fluid biomarkers and

rate of cognitive decline in very mild dementia of the

Alzheimer type.Arch Neurol 2009;66:638–45.

47 Donohue MC,Gamst AC,Thomas RG,Xu R,Beckett L,

Petersen RC,Weiner MW,Aisen P,Alzheimer’s Disease

M.N.Samtani et al.

160/

75:1/ Br J Clin Pharmacol