Opportuni)es  and  Challenges  in  Drug  Discovery  for  Au)sm  Spectrum  Disorders  

Lawrence W Fitzgerald, PhD, MBA 03/03/15

1

Autism – Spectrum of disability with significant unmet medical need

Social Communication Deficits

Fixed interests and repetitive behaviors

Aggression/ irritability

Anxiety

Intellectual disability

Inattention Hyperactivity

Epilepsy

Sleep disorder

Core symptoms Adjunct symptoms Comorbid illness

Risperidone Abilify (FDA approved)

SSRIs (off-label)

Stimulants (off-label)

2

Location N Age (y) M:F Period Prevalence

Brick Township, NJ1 17,792 3-10 2.7:1 1998 0.67% (1/149)

Metro Atlanta2 289,456 3-10 3.8:1 1996 0.34% (1/294)

NHIS, US National3 18,885 4-17 3.7:1 2003-4 0.57% (1/175)

NSCH, US National4 79,590 4-17 3.7:1 2003-4 0.55% (1/182)

CambridgeshireUK5 NA 5-11 4:1 1999 0.57% (1/175)

South Thames, UK6 56,946 9-10 3.3:1 1999-2001 1.16% (1/86)

CDC ADDM Network7 363,749 8 4.5:1 2010 1.47% (1/68)

1. JAMA 2001, 286:3093-99 1. Pediatrics 2001, 108:1155-61 2. JAMA 2003, 289:49-55 3. MMWR 2006 55:481-6 4. MMWR 2006 55:481-6 5. Autism 2002, 6:331-7 6. Lancet 2006, 210-5 7. ADDM Network Report, CDC 2014

Autism Prevalence (US/UK)

3

Autism prevalence – why the trending?

§  Estimates regarding the prevalence of autism (and later ASD) have increased exponentially –  For decades, the prevalence was thought to be 4 to 5 cases per

10,000 children in the US

–  In a 2014 report, CDC network study (covering 2010) resulted in a revised estimate of approximately 1 child in every 68 in the US will have an ASD

–  Approximately 50% will have significant to borderline intellectual disability where as 50% will have average to above average IQ

§  Increased prevalence is likely driven by publicity, reduction in stigma, broadening/defining diagnosis (e.g., ID→ASD), public services

4

ASD Genetics

§  Autism is among the most heritable of all neuropsychiatric disorders…some of genetic changes are spontaneous!

§  Concordance rates –  82-92% for monozygotic twins; 1-10% for dizygotic twins

§  Multiple underlying genes –  complex disorder, phenotypic variability

–  chromosomal abnormalities

–  CNVs - copy number changes (duplication/deletions; e.g., 15q11, 22q13),

–  rare single gene mutations - Monogenic syndromic ASDs (FXS, Rett, Angelmans, TSC) and others (Shank3, Pten)

5

Life time costs of autism

6

Ganz, M. L. Arch Pediatr Adolesc Med 2007;161:343-349. Buescher et al., JAMA Pediatr. 2014:721-728.

Age distribution of incremental per capita societal costs of autism

*                    *  

§   Life-­‐long  costs  of  ASD  o  $2.4  Million  when  au)sm  involves  ID  o  $1.4  Million  when  ID  not  present  

§  In  childhood,  majority  costs  are  special  educa)on  services  and  loss  of  income/produc)vity  of  parents  

§  In  adult  years,  it  is  mostly  comprised  of  residen)al  and  suppor)ve  services  and  loss  produc)vity/employment  of  individual  

7

Neurologic, Psychiatric, and Ophthalmologic disorders comprise one of the largest global pharmaceutical markets accounting for $121 billion in sales in 2009

                           

7  

Neuroscience has a proven commercial track record

Serendipity  has  enabled  CNS  (Psychiatric)  market  success  

Drug   Original  use   Psychiatric  use   FDA  approval  

Amphetamine   Bronchodilator   AQen)on,  Obesity   1937  

Carbamazepine   Epilepsy   Bipolar  Mania   2004  

Chlordiazepoxide   Chemical  dyes   Anxiety   1960  

Chlorpromazine   Surgical  relaxa)on   Psychosis   1954  

Isoniazid   Tuberculosis   Depression   1958  

Lithium   Gout   Mania/BP   1970,  1974  

Meprobamate   Preserva)ve   Anxiety/tranquilizer   1950  

Reserpine   An)hypertensive   Psychosis   1954  

Valproate   Epilepsy   Bipolar  disorder   1995  

Ketamine   Anesthesia   Tx-­‐resistant  Depression   ????  

It  is  notable  for  ASD,  despite  its  incidence,  that  it  has  not  benefited  from  serendipity!  8

…..followed  by  the  emergence  of  target  and  structure-­‐based  based  drug  discovery  

Diphenhydramine    (Benedryl)  

An)depressant    ac)vity  

Monoamine    Uptake  Blockade  

SSRIs:  Prozac,    Zoloe,  Paxil  NRIs:  StraQera  SNRI:  Cymbalta  NSRI:  Effexor      

1946   1970   1972   1987+  

dibenzazepine    (TCA  an)depressants)  

1966  

clozapine    (first  atypical  an)psycho)c)  

1971  

clozapine    (reintroduced  for    Tx-­‐resistant  psychosis);  Pharmacology  studied  

1989  Zyprexa  Risperdal  Geodon  Seroquel  Abilify      

1990-­‐2000’s  

Pharmacology-­‐inspired  

Examples  of  new  psychiatric  medicines  that  emerged  prospecQvely  from  a    modern  model  of  disease  are  quite  lacking!  -­‐-­‐-­‐-­‐  since  2007,  no  new  MOA’s  approved  by  FDA  only  reuptake  blockers  and  atypical  APDs  

9

Medchem  structure-­‐inspired  

Transition from pharmacology models to disease mechanisms has been difficult – case example, antidepressants

10  

Target  ID/valida)on  

POM/POC  Phase  1/2  

PredicQon!!  

Benadryl  

Prozac  

Effexor  Zoloe  

Paxil  

Experimental  Tests:  1)  Measure  5-­‐HT/NE/

DE  uptake  2)  Specific  In  vivo  

rodent  tests  

Reuptake  blockers  (1980-­‐2000’s)   Blockbuster  AnQdepressants  

PredicQon/TranslaQon  

New  target  mechanism  (1990’s  to  present)  1)  CRF  2)  Vasopressin  3)  5-­‐HT2  4)  Substance  P  

Experimental  Tests:  1)  Shie  to  animal/human  

transla)on  2)  Modeling  disease  

mechanisms  -­‐  neurobiology,  behavior,  genomics  

1980-­‐2000s  

2000-­‐present  

10

11  

Universities, Gov’t

Knowledge  generaQng  centers  

Private R&D $100B+ NIH $31B

Corporate R&D, NGO Gov’t, Venture $$

Sources  of  capital  

Product/value generating centers

Pharma, Biotech

+

+

=

Drug discovery ecosystem shows a decline in innovation & growing neglect of key disease areas and markets

Why  ?  1980’s                                    2010’s  o  Poor  research  model  predic)on/transla)on  o  Knowledge  gaps  in  biology/medicine  o  Soe  clinical  endpoints,  pa)ent  heterogeneity  o  Inadequate  scien)fic  reliability/reproducibility  o  Business  reasons  

o  Organiza)onal  turmoil  (M&A/restructuring)  o  Risk  aversion  o  Misaligned  incen)ves  o  Neglect  markets  –  CNS,  An)bio)cs  

ü  80%  world’s  pa)ents  w/o  insurance  

Products, Customer value

Drug  discovery  process  and  risk  mi)ga)ng/value  crea)ng  milestones  

Screening  strategy  

Hit/Lead  ID  

Lead  Op)miza)on/Candidate  ID  

Target  ID/valida)on  

Preclinical  Dev  

Phase  1  Dev  

   

Rela)ve  Risk  

$$  Invested  and  Value  Created  

Phase  2/3/registra)on  

Mechanis)c  link  to  human  

disease  IP  Genera)on  and  FTO  

Molecules  with  Drug-­‐like  proper)es;  

Target  Candidate  Profile  

Adequate  safety  margin  for  human  tes)ng  

Adequate  PK,  Safety  and  Efficacy  in  Humans  

Target  Product  Profile  

12

Right  target?                                  Right  drug/molecule?            Right  pa)ent?         Right  test?  

Bridging  the  knowledge  gap  –  is  it  Au)sm’s  )me?  

13  

R&D strategic opportunities •  Tx  of  adjunct/comorbid  illness  (e.g.,  irritability,  ID,  ADHD,  sleep,  mood)  

–  Ideal  candidates  do  not  exacerbate  core  symptoms  –  Consider  combining  w  behavioral  interven)ons  to  show  benefits  –  Upside:  “winning  with  low  hanging  fruit”,  Downside:  low  commercial  appeal/no  

differenQated  benefit    

•  Anchor  to  rare  molecular  variant,  sequenQally  expand  to  other  mechanisQcally  convergent  (similar)  disease  segments    –  U)lize  genomics,  systems  and  mechanis)c  biology,  translatable  biomarkers,  pa)ents  

stra)fica)on  strategies.  Examples,  prolifera)ve  pathways,  I/E  balance,  protein  synthesis  –  Upside:  “Targeted  medicine,  strong  mechanisQc  link  to  eQology  of  disease;  Downside:  

Efficacy  limited  to  rare  and  narrow  segment,  potenQally  expensive  drugs    

•  UQlized  systems  biology/neurochemical/behavioral  approach  to  tx  symptoms  –  Modulators  of  specific  neural  circuits  that  underlie  symptoms  domains  (e.g.,  OT)  –  U)lize  gene)c  models  and  natural  rodent  strains  with  symptoms  that  pass  face  validity  –  Upside:  “broader  appeal”,  tx  diverse  forms  of  ASD”  Downside:  there  is  no  magic  bullet  

14

Confiden)al  |  Copyright  2011  SAGE  Therapeu)cs   15  

Pfizer Internal Use

Risperdal ASD Program- limited clinical trial requirements for FDA approval

§  Two, 8 week, placebo-controlled trials in children 5-16 who met DMS-IV criteria for autistic disorder (n=101, n=55)

§  Primary outcome measure in both studies was the change from baseline to endpoint in the irritability subscale of the ABC (ABC-I).

15

Key mechanistic frameworks in etiology of ASD

16  

       The  balance  of  synapse  forma)on  /  matura)on  /  elimina)on  How  to  maintain  E/I  balance  à  op)mal  fitness?    Balance  can  be  disrupted  by  gene-­‐defects  regula)ng  protein  synthesis!  

Kelleher  and  Bear,  Cell,  2008.  

Candidate drug target of the synapse

pre  

post  

Keith  and  El-­‐Husseini,  2008   17

Top–down: Integrative and translational approach to target validation and drug discovery

18  

Genomic analysis

Identification of risk gene (e.g., FMR1)

Define variation of sequence of risk gene (loss, gain, altered function?)

Determine biologic influence of gene variant in cells or network of cells (e.g., primary cells, iPSCs)

Determine genetic effect on in vivo phenotype in rodents (Mouse model)

Confirm altered function/expression of gene and associated pathways in humans

maternal paternal

Target ID

Target validation

Bottom up: Path for building molecular models of disease for ASD from gene variant to validated target

19

Macrocephaly in early ASD and recapitulated in mice with conditional KO of Pten

Brain Size Difference in Human ASD by Age

Progressive Macrocephaly and Regional Hypertrophy in the Pten Mutant Mice

Courchesne et al., Neuron, 2007 Kwon et al, Neuron, 2006 20

STX209  (r-­‐baclofen)  corrects  synapQc  phenotypes  in  fmr1  ko  mice  

21 Henderson et al, 2012

R-­‐bac  reduces  spine  density   R-­‐bac  reverses  excess  protein  synthesis  

R-­‐bac  reverses  excess  AMPA  receptor  internalizaQon  

22

Walsh and Hochedlinger, Cell, 2010, 143:499-500

Marchetto et al., Cell, 2010, 143:527-539

A model for neural development and treatment of Rett Syndrome using human induced pluripotent stem cells

1 2

3

FMR1

pten

Shank3

TSC1/2

Analysis of GeM mice or human iPSC Common mechanisms underlying behavioral & physiological phenotypes

Bridge to ASD patient segments or endophenotypes with biomarkers, clinical features?

Best targets for aggregating segments of patients

From discrete genetic association to more common targets and pathways

Q: Can the rare monogenic forms of ASDs inform how to tx larger segments of patients?

23

pen  pen  

NRL4  FMR1   TSC1/2 FMR1 ?

pten  

Shank3

TranslaQonal  strategy  

Category and utility of biomarkers Biomarker is a laboratory measurement or physiological sign in association with a physiological process of therapeutic or diagnostic value

§  Target biomarker – measures physical or biological interaction with the molecular target (PET ligand demonstration of receptor occupancy, measurement of enzyme inhibition); o  benefits: PK/PD (dose selection), demonstrate target engagement o  PET ligand, e.g., mGlur5 NAM; CSF levels of drug

§  Mechanistic biomarker – measures a biological effect presumed to be downstream of molecular target

o  benefits: like target + target-mediated activity,

o  may be change in physiological (blood flow), biochemical (substrate turnover), behavioural (reaction time), genetic (gene expression) or proteomic (composition of protein in tissues/biofluids).

o  qEEG, CSF/plasma markers §  Outcome/surrogate marker – impact on pathophysiology that links to clinical

efficacy. Opportunity for patient segmentation, predict long-term efficacy o  protein synthesis (11C-leucine PET), ERP –EEG, fMRI/DTI, eye gazing etc.

24  

25  

Schultz,Int  J  Dev  Neurosci  2005  

Interest  in  people  (face  detecQon)  

Alending  to  people  (face  idenQficaQon)  

Thinking  and  learning    about  people  (emoQonal  processing)    

Neural systems involved in the development of social expertise in humans

26  

ERP/face  recogniQon   Eye  tracking  

fMRI  -­‐  amygdala  and  fusiform  gyrus  

Provided  by  Yale  Child  Study  Center  

Using outcome & mechanistic biomarkers to bridge the gap from target validation to testing efficacy in humans

27  

Key clinical questions for clinical protocol design and testing a new therapeutic

March  5,  2015  Confiden)al  |  Copyright  2011  SAGE  Therapeu)cs   27  

Developmental Time (t)

Beh

avio

ral

Cap

acit

y

Typical

FXS

FXS + DM

FXS + Symp

Critical Period

Post- Critical Period

1)  MOA  -­‐  Symptoma)c  or  disease  modifying  agent?  

2)  Ideal  tx  period,  subject  age  and  length  of  trial?  

3)  Pa)ent  selec)on,  e.g.,  severity  of  symptoms,  presence  of  ID,  or  other  comorbid  illness?  

4)  What  is  your  primary  outcome  measure?    

5)  Condi)ons  of  doing  the  subject  ra)ngs?  

6)  Concerns  re:  placebo  effect?  7)  Any  objec)ve,  validated  

biomarkers  for  proof  of  mechanism,  or  proof  of  concept?  

28  

Validation of ABC-C for Fragile X

ORIGINAL PAPER

Psychometric Study of the Aberrant Behavior Checklist in FragileX Syndrome and Implications for Targeted Treatment

Stephanie M. Sansone • Keith F. Widaman • Scott S. Hall • Allan L. Reiss •

Amy Lightbody • Walter E. Kaufmann • Elizabeth Berry-Kravis •

Ave Lachiewicz • Elaine C. Brown • David Hessl

Published online: 5 October 2011! Springer Science+Business Media, LLC 2011

Abstract Animal studies elucidating the neurobiology offragile X syndrome (FXS) have led to multiple controlled

trials in humans, with the Aberrant Behavior Checklist-

Community (ABC-C) commonly adopted as a primaryoutcome measure. A multi-site collaboration examined the

psychometric properties of the ABC-C in 630 individuals

(ages 3–25) with FXS using exploratory and confirmatoryfactor analysis. Results support a six-factor structure, with

one factor unchanged (Inappropriate Speech), four modi-

fied (Irritability, Hyperactivity, Lethargy/Withdrawal, andStereotypy), and a new Social Avoidance factor. A com-

parison with ABC-C data from individuals with general

intellectual disability and a list of commonly endorseditems are also reported. Reformulated ABC-C scores based

on this FXS-specific factor structure may provide added

outcome measure specificity and sensitivity in FXS clinicaltrials.

Keywords FMR1 gene ! Fragile X syndrome ! Autism !Factor analysis ! Rating scale ! Social avoidance

Fragile X syndrome (FXS) is the leading known cause of

both inherited intellectual disability (ID) and autism, andrecent estimates of the frequency of the full-mutation are as

high as one in every 2,500 births (Hagerman 2008). FXS is

caused by the expansion of a trinucleotide repeat sequence,cytosine-guanine-guanine (CGG), in the promoter region

of the Fragile X Mental Retardation 1 gene (FMR1) on the

long arm of the X chromosome at Xq27.3. Expansions

A portion of the data was presented at the 44th Annual GatlinburgConference, San Antonio, TX, in March 2010.

Electronic supplementary material The online version of thisarticle (doi:10.1007/s10803-011-1370-2) contains supplementarymaterial, which is available to authorized users.

S. M. Sansone ! D. Hessl (&)Medical Investigation of Neurodevelopmental Disorders(MIND) Institute, University of California Davis MedicalCenter, 2825 50th Street, Sacramento, CA 95817, USAe-mail: david.hessl@ucdmc.ucdavis.edu

K. F. WidamanDepartment of Psychology, University of California, Davis,One Shields Avenue, Davis, CA, USA

S. S. Hall ! A. L. Reiss ! A. LightbodyDepartment of Psychiatry and Behavioral Sciences, Center forInterdisciplinary Brain Sciences Research, Stanford University,401 Quarry Road, Stanford, CA, USA

W. E. KaufmannCenter for Genetic Disorders of Cognition and Behavior,Kennedy Krieger Institute, Johns Hopkins University Schoolof Medicine, Baltimore, MD, USA

E. Berry-KravisDepartments of Pediatrics, Neurological Sciences, Biochemistry,Rush University Medical Center, Chicago, IL, USA

A. LachiewiczDepartments of Pediatrics and Psychiatry and BehavioralSciences, Duke University Medical Center, Durham, NC, USA

A. Lachiewicz ! E. C. BrownPrivate Practice, Reno, NV, USA

D. HesslDepartment of Psychiatry and Behavioral Sciences,University of California, Davis, School of Medicine,Sacramento, CA, USA

123

J Autism Dev Disord (2012) 42:1377–1392

DOI 10.1007/s10803-011-1370-2

•  ABC-C scale originally validated in the MR/ID population •  Multi-site collaboration examined the psychometric properties for ABC in 630 individuals with FXS, ages 3-25 yrs •  Results supported a 6 factor scale structure with one unchanged (inappropriate speech), 4 modified scales (Irritability, Hyperactivity, Lethargy/withdrawal, Stereotypy), and the additional of one new factor, Social Avoidance •  The new reformulated ABC-C scores based on FXS-specific factor structure may provide outcome measure specificity and sensitivity for FXS clinical trials

Confiden)al  |  Copyright  2011  SAGE  Therapeu)cs   29  

tant physiological role (23), is the main structure of reward expectationand placebo response (24).

Here, methylation of the FMR1 promoter in whole blood was as-sessed with an MSP assay of bisulfite-treated DNA and sequencing ofbisulfite-treated DNA clones. The sensitivity of these techniques is highcompared with the standard methylation-specific Southern blot. Thisdifference in techniquesmay explain the relatively high observed occur-rence of partial FMR1 promoter methylation in a population in whichall of the patients carry the fullmutation. In addition, we cannot excludeascertainment bias toward higher-functioning patients, and thus alsotoward individuals with partial FMR1 promoter methylation, giventhe rigorous study schedule and the procedures involved. The detectionof partial methylation at the FMR1 promoter may suggest the presenceof a mosaic of premutation (between 55 and 200 CGG repeats withoutmethylation at the FMR1 promoter and usually few to no symptoms ofFXS) and full-mutation (more than 200 CGG repeats) alleles. Becauselarge premutations (100 to 200 CGG repeats) have been associated withabove-normal levels ofFMR1mRNA,but reducedFMRP(21), thenormallevels of FMR1 mRNA detected in two patients could be explained by amosaic of premutation and full-mutation alleles. Partial FMR1 promotermethylation could also be explained by distinct FMR1 promoter meth-ylation status in different tissues after somatic expansion of the CGGrepeats in the early stages of embryonic development. Investigation intothe tissue-specific pattern ofFMR1 promotermethylation and expressionmay shedmore light on the functional consequences of FMR1 dysregula-tion on disease phenotypes and drug response and may offer additionalopportunities for discovering other indicators of treatment response.

AFQ056 was reasonably well tolerated in this population, with allpatients completing the up-titration, reaching the target dosage, andcompleting the AFQ056 period of treatment. Adverse events were re-ported by 80%of the patients in the study,with these eventsmainlymildor moderate in severity, andmost commonly fatigue and headache. Fa-tigue was reported in both AFQ056 and placebo treatment groups,whereas headachewas only reported dur-ing AFQ056 treatment. Only one patientexperienced a serious adverse event, andbecause this occurred while he was in theplacebo treatment period, the event wasnot suspected to be related to the studydrug.

This study was designed to investigatethe effects of AFQ056 treatment on aber-rant behaviors in patients with FXS andthe safety and tolerability of this drug inthis population. A two-treatment, two-period, crossover design was used to al-low intraindividual comparisons andhencereduce the number of patients required toachieve satisfactory statistical power. To re-duce interpatient variability, we recruitedonly men with the full FXS mutation tothe study. These men were aged between18 and 35 years, because AFQ056 pharma-cokinetic data are only available for adultsand because there are few available data onFXS patients older than 35 years. The treat-ment regimen in this study used AFQ056doses shown to have an acceptable safety

and tolerability profile in healthy subjects. Because this was the firstexperience of AFQ056 in patients with FXS, a slow titration dosingregimen was used. Because AFQ056 plasma concentrations were ex-pected to reach a steady state within 3 to 4 days, 4-day intervals betweendose escalations were considered sufficient to assess the safety and tol-erability of each dose. The 28-day treatment periods were chosen on thebasis of the available AFQ056 toxicology data, and the 1-week washoutperiod was considered sufficient to avoid carryover effects becausecomplete elimination of AFQ056 was expected after ~3.5 days.

This study is limited by the small sample size of 30 patients, thecrossover design, and the short time scale. This is particularly relevantfor the exploratory analysis comparing patients with full and partialFMR1 promoter methylation, where only a small number of patientswith full FMR1 promoter methylation were shown to respond signifi-cantly to AFQ056 treatment. The levels of FMR1mRNAwere assessed,but assessment of the levels of FMRP in each patient may have aidedanalysis of the results. It seems likely that patientswith fullmethylation atthe FMR1 promoter and no detectable FMR1 mRNA do not expressFMRP, but knowing the levels of FMRP expression in patientswith partialFMR1 promoter methylation may have helped explain the variable treat-ment response in these patients. Larger, longer-term, prospective trials arerequired to confirm the findings of this study, to test the long-termefficacyof AFQ056 on behavior, and to study any effects on the cognitive deficitsassociated with FXS. If these results are positive, then future studies will beneeded to test safety and efficacy in childrenwithFXS, because the greatestbenefits from mGluR5 inhibition may be derived during development.

We have shown that the selective inhibition of mGluR5 by AFQ056can provide a significant effect on behavioral problems in a subpop-ulation of patients with FXS. If confirmed in future studies, these resultssuggest that AFQ056 may provide valuable improvement in behavioralsymptoms of FXS and that this improvement is predicted by full meth-ylation at the FMR1 promoter. Why subjects lacking FMR1 expressionare better responders to the AFQ056 treatment remains an open ques-

60

A

40

20

0

Chan

ge fr

om b

asel

ine

to d

ay 1

9/20

on

ABC

–C

–20

–40

–60

60

B

40

20

0

–20

–40

–60Placebo AFQ056 Placebo AFQ056

Fig. 3. (A andB) A comparison of the effect of AFQ056 andplacebo treatments on the change frombaselineto day 19 or 20 on the ABC-C score in individual patients with (A) full methylation at the FMR1 promoter and(B) partial methylation at the FMR1 promoter. A decrease in ABC-C score indicates an improvement in be-havioral symptoms.

R E S EARCH ART I C L E

www.ScienceTranslationalMedicine.org 5 January 2011 Vol 3 Issue 64 64ra1 6

on

Sep

tem

ber

14, 2

012

stm

.sci

ence

mag

.org

Dow

nloa

ded

from

FRAG I L E X SYNDROME

Epigenetic Modification of the FMR1 Gene in Fragile XSyndrome Is Associated with Differential Responseto the mGluR5 Antagonist AFQ056Sébastien Jacquemont,1* Aurore Curie,2* Vincent des Portes,2 Maria Giulia Torrioli,3

Elizabeth Berry-Kravis,4 Randi J. Hagerman,5 Feliciano J. Ramos,6 Kim Cornish,7 Yunsheng He,8

Charles Paulding,8 Giovanni Neri,9 Fei Chen,1,10 Nouchine Hadjikhani,10,11 Danielle Martinet,1

Joanne Meyer,8 Jacques S. Beckmann,1 Karine Delange,2 Amandine Brun,2 Gerald Bussy,2

Fabrizio Gasparini,12 Talita Hilse,13 Annette Floesser,13 Janice Branson,12 Graeme Bilbe,12

Donald Johns,14 Baltazar Gomez-Mancilla14†

Fragile X syndrome (FXS) is an X-linked condition associated with intellectual disability and behavioral problems. It iscaused by expansion of a CGG repeat in the 5! untranslated region of the fragile X mental retardation 1 (FMR1) gene.Thismutation is associatedwith hypermethylation at the FMR1promoter and resultant transcriptional silencing. FMR1silencinghasmany consequences, includingup-regulation ofmetabotropic glutamate receptor 5 (mGluR5)–mediatedsignaling. mGluR5 receptor antagonists have shown promise in preclinical FXS models and in one small open-labelstudy of FXS. We examined whether a receptor subtype–selective inhibitor of mGluR5, AFQ056, improves thebehavioral symptoms of FXS in a randomized, double-blind, two-treatment, two-period, crossover study of 30 maleFXS patients aged 18 to 35 years. We detected no significant effects of treatment on the primary outcome measure,the Aberrant Behavior Checklist–Community Edition (ABC-C) score, at day 19 or 20 of treatment. In an exploratoryanalysis, however, seven patients with full FMR1 promoter methylation and no detectable FMR1 messenger RNAimproved, as measured with the ABC-C, significantly more after AFQ056 treatment than with placebo (P < 0.001). Wedetected no response in 18 patients with partial promotermethylation. Twenty-four patients experienced an adverseevent, which was mostly mild to moderately severe fatigue or headache. If confirmed in larger and longer-termstudies, these results suggest that blockade of the mGluR5 receptor in patients with full methylation at the FMR1promoter may show improvement in the behavioral attributes of FXS.

INTRODUCTION

Fragile X syndrome (FXS) is an X-linked genetic condition associatedwith intellectual disability and behavioral problems including anxiety,aggression, hyperactivity, impulsivity, shyness, attention deficit dis-order, and autism (1). It is caused by expansion of a CGG trinucleotiderepeat in the 5! untranslated region of the fragile X mental retardation

1 (FMR1) gene. This mutation is associated with hypermethylation atthe FMR1 promoter and consequent transcriptional silencing (2–5).The FMR1 protein (FMRP) is a cytoplasmic RNA binding proteinknown to repress the translation of messenger RNAs (mRNAs) atsynapses (6). It has been suggested that in the absence of FMRP, lossof repression ofmetabotropic glutamate receptor 5 (mGluR5)–mediatedpathways results in the behavioral and cognitive impairments associatedwith FXS (7).

Recent studies in animal models of FXS have suggested that manyaspects of the FXS phenotype, including behavioral abnormalities, cog-nitive deficits, and altered dendritic spines, may be attributable to exces-sive signaling by mGluR5, a group I mGluR. Genetic down-regulationof mGluR5 expression by crossing Fmr1 knockout mice with heterozy-gous Grm5 knockout mice rescues many of the FXS phenotypes, withthe exception of macroorchidism (8). Consequently, selective mGluR5antagonists may offer effective treatment for the symptoms of FXS.These compounds have been available as research tools for nearly adecade, and the prototype of the class, 2-methyl-6-(phenylethynyl)-pyridine (MPEP), has been tested extensively in animal models ofFXS.MPEP rescues themost robust central nervous system phenotypesin Fmr1 knockout mice, namely, hyperactivity and audiogenic seizuresusceptibility (9). Cognitive and neuroanatomical phenotypes were res-cued in a fruit fly model (10), neurite branching and craniofacial ab-normalities were rescued in a zebrafish model (11).

1Service de Génétique Médicale, Centre Hospitalier Universitaire Vaudois, CH-1011Lausanne, Switzerland. 2National Reference Center for Fragile X and Other XLMR, HospicesCivils de Lyon, Université de Lyon and CNRS UMR 5230 (L2C2), F-69675 Bron, France.3Università Cattolica del Sacro Cuore, Cattedra di Neuropsichiatria Infantile, 00168 Rome,Italy. 4Departments of Pediatrics, Neurological Sciences, and Biochemistry, Rush UniversityMedical Center, Chicago, IL 60612, USA. 5Department of Pediatrics and MedicalInvestigation of Neurodevelopmental Disorders (M.I.N.D.) Institute, University of California,Davis, School of Medicine, Sacramento, CA 95817, USA. 6Department of Pediatrics, Universityof Zaragoza Medical School, 50009 Zaragoza, Spain. 7Centre for Developmental Psychiatryand Psychology, School of Psychology and Psychiatry, Monash University, Melbourne,Victoria 3800, Australia. 8Biomarker Development, Novartis Institutes for BiomedicalResearch, Cambridge, MA 02139, USA. 9Università Cattolica del Sacro Cuore, Istituto diGeneticaMedica, 00191 Rome, Italy. 10BrainMind Institute, École Polytechnique Fédérale deLausanne, CH-1015 Lausanne, Switzerland. 11Martinos Center for Biomedical Imaging,Harvard Medical School, Boston, MA 02129, USA. 12Neuroscience Discovery, NovartisPharma AG, CH-4056 Basel, Switzerland. 13Neuroscience Clinical Sciences and Innovation,Novartis Institutes for Biomedical Research, Novartis PharmaAG, CH-4056 Basel, Switzerland.14Neuroscience Translational Medicine, Novartis Institutes for Biomedical Research, NovartisPharma AG, CH-4056 Basel, Switzerland.*These authors contributed equally to this work.†To whom correspondence should be addressed. E-mail: baltazar.gomezmancilla@novartis.com

R E S EARCH ART I C L E

www.ScienceTranslationalMedicine.org 5 January 2011 Vol 3 Issue 64 64ra1 1

on

Sep

tem

ber

14, 2

012

stm

.sci

ence

mag

.org

Dow

nloa

ded

from

Science Translational Medicine, 2011 29  

Phase 2 results of Arbaclofen in FXS: post-hoc analysis on social withdrawal

30  

FXS Study 22001: Post-hoc analyses

(n=27) STX209

(mean ± SD)!Placebo

(mean ± SD)! p-value!

ABC-Social Withdrawal -4.3 ± 6.3 -0.4 ± 7.1 < 0.05

Vineland – Socialization (raw) 14.2 ± 19.0 4.6 ± 10.8 < 0.05

CGI-I 2.7 ± 1.1 3.5 ± 1.2 < 0.01

CGI-S -1.0 ± 1.1 -0.3 ± 0.9 = 0.01

Treatment preference (clinician) 63% 19% < 0.01

Treatment preference (parent) 67% 19% = 0.001

Responders (CGI-I =1 or 2, and ABC-SW improvement ≥ 25%) 42% 7% < 0.01

ABC-Social Avoidance -2.1 ± 2.9 0.04 ± 2.9 = 0.011

10

Subjects with ABC-Social Withdrawal > 8 at screening, baseline1 & baseline2

* Provided by Paul Wang MD, Seaside Therapeutics, GRC meeting June 2012

Mi)ga)ng  Phase  2  efficacy  (PoC)  risk  in  FXS  (Seaside,  Roche  and  Novar)s)  

Loss  of  func)on  of  FMR1  

Increase  dendri)c  protein  

Clinically  relevant  spine  morphology?  (spine  density,  immature  

spines)  

Target-­‐linked  pathophysiology?  (AMPA  receptors/LTD)    

 Clinically  Relevant    

behavioral  phenotypes?  

(seizure,  arousal,  anxiety?  Cogni)on?  

   

Pa)ent  entry  criteria/

segmenta)on  

Trial  design  and  outcome  measures  

Outcome  surrogates  links  target  with  correc)on  of  

pathology  and  offers  segmenta)on  

Target/mechanis)c  biomarker  for  

demonstra)ng  target  engagement  and  picking  doses  

                 TI/TV  milestone  Candidate  selecQon              Proof  of  Mechanism          Proof  of  Concept  

                                 Proof  of  Principle    Preclinical                    Phase  1B        Phase  2A                        

31

32  

*ABC-­‐C  showed  no  sta)s)cally  meaningful  effect…only  their  ReQ-­‐specific  clinical  scale  

Our past and present: challenges of science business

§  Biotech/ Pharma model is evolving

-  Drug discovery is hampered by persistent risk and uncertainty

-  Emergence of virtual biotech/pharma unit with flexible/low cost structure

-  Emergence of flexible risk sharing models of public/ private partnership (pharma/bio + gov’t/academia, consortia) -  Shifting drug discovery risk to VC funded enterprise, partnering at value inflection point (phase 2 and later).

Pharm Bio Pharm

Acad/ gov’t

Pharm Bio

product product

Bio product

First Gen (1980’s) Second Gen (1990’s) Current Gen (2000’s)

33

Nonprofit pharma like Mend Therapeutics, a new business model for innovation

34  

MEND is a nonprofit pharmaceutical company that seeks to combat diabetes and its complications globally by promoting prevention and early detection, expanding access to quality medical care, and developing life-changing therapies

35  

Well  Child  Lens  Au)sm  Educate  and  facilitate  

communicaQon  between  all  healthcare  stakeholders  

• www.wellchildlens.com:  more  than  600  videos  on  the  symptoms,  diagnosis,  and  treatment  of  ASD,  for  parents  and  pediatric  HCPs.  

• The  Well  Line  App:  interac)ve  mobile  app,  tracks  a  child’s  development  birth  to  age  3,    key  developmental  milestones,  connects  parents  to  the  pediatric  HCP,  reminders  about  screenings.  

Drive  efficient,  proper  and  Qmely  diagnosis  

• The  InteracQve  Modified  Checklist  for  AuQsm  in  Toddlers  (iM-­‐CHAT):    (for  children  16  –  30  months)  digital,  video-­‐enhanced  version  of  the  M-­‐CHAT,  with  instant  follow-­‐up,  scoring  and  results.  

• Well  Visit  Video  Guide:  strategies  to  prompt  cri)cal  behaviors  in  toddlers  during  a  well  visit.  

• CME  Courses:  3  accredited  courses  

Enhance  office  workflow  opQmizaQon  to  save  Qme  and  increase  revenues  

• EMR  CompaQble  Clinical  Version:  • Dashboards  for  each  doctor  • Automa)c  screening  reminders  • Results  in    the  pa)ent  record  • Streamlines  the  3-­‐step  process  for  screening  and  surveillance  

• Saves  staff  )me  and  resources  • Automa)c  codes,  search  by  specific  insurance  plans,  ensure  reimbursements  and  increase  revenues  

Summary  

36

•  Develop  pathophysiological  based  models  and  measure  mulQple  efficacy  endpoints  that  are  relevant  to  symptoms  and  potenQal  translatable  in  the  clinical  sepng  o  Build  quality,  detailed,  integrated  data  sets  at  cellular,  circuit,  and  animal  level  o  Iterate  on  past  findings  to  understand  which  observa)on  are  robust  and  reproducible  o  Focus  on  endpoints  that  have  construct  not  face  validity    

•  Determine  points  of  convergence  for  geneQc  models  of  ASD  to  cluster  disparate  forms  of  ASD  and  provide  common  points  of  intervenQon  for  novel  therapeuQcs  o  Possibili)es  include  convergence  inferred  at  the  level  of  circuit,  cell  func)on/signaling,  

genome/transcriptome  o  Points  of  convergence  will  likely  be  symptoms  specific…  magic  bullets  are  rare  o  How  do  we  reduce  this  to  prac)ce,  re:  pa)ent  segmenta)on  strategies  in  the  clinic    

TranslaQonal  Science  Planning  •  Develop  mechanisQc  and  target  biomarker  for  demonstraQng  target  engagement,  

understanding  PK/PD  relaQonship  and  projecQng  human  dose  •  Develop  sensiQve  efficacy/outcome  biomarkers  starQng  with  animal  modeling  

o  Confirm  impact  on  disease  mechanism  prior  to  registra)on  trials  o  Provide  an  opportunity  to  stra)fy  or  enrich  pa)ents  groups  that  will  respond  robustly  o  Be  conscious  of  cost  and  rela)ve  ease  to  validate  in  clinical  sexng  

Target  ID/validaQon  

Summary  –  con’t  

37  

•  Need  to  develop  opQmal  clinical  design  elements  for  detecQng  efficacy  of  therapeuQcs  on  core  symptoms  of  ASD  o  How  to  segment  pa)ents,  manage  pa)ent  heterogeneity  (e)ology/age/ID,  severity  of  

symptoms  etc),  manage  clinical  trial  error/variance,  (e.g.,  rater  bias  and  error),  establish  ideal  primary  outcome  measures  and  QOL  instruments?  

o  We  need  to  deal  with  our  past,  problem  of  robust  placebo  effects  in  psychotherapeu)cs;  consider  managing  expectancy,  use  trial  design  like  placebo  run-­‐in,  ID  biomarkers  

o  Let’s  iden)fy  more  objec)ve,  measurable  outcome  biomarker/surrogates  for  efficacy    Discovery/development  partnerships  •  Form  strong  mulQdisciplinary  teams  between  private  and  public  preclinical  and  clinical  

scienQsts  to  guide  preclinical  and  clinical  translaQonal  research  o  U)lize  cross-­‐ins)tu)onal,  public/private  funding  mechanisms  (NCAT,  U  grants,  NIH  Blueprint)  o  Encourage  greater  investment  by  NGOs  in  private  discovery/development  programs  

Other  InnovaQons  o  New  business  models,  e.g.,  sustainable,  product  focused  nonprofit  pharma  o  Consider  and  study  behavioral  therapy/drug  combina)ons  or  synergy  o  Health  IT  solu)ons  that  improve  symptom/QoL,  or  drive  earlier  diagnosis  

Clinical  trials  

Acknowledgements

38  

Pfizer  AuQsm  Unit  (co-­‐founders/partners/advisors)  Diane  Stephenson  (Cri)cal  Path)  Howard  Mayer  (Shire  HGT)  Mahesh  Lachyankar  (MEND  Therapeu)cs)  Mat  Pletcher  (Pfizer  Rare  Diseases)  Mark  Bear  (MIT,  Seaside  Tx)  Michael  Ehlers  (Pfizer)  Rob  Malenka  (Stanford)    Nonprofits  Rob  Ring  (Au)sm  Speaks)  Dan  Smith  (Au)sm  Speaks)  Steve  Roberds  (Tuberous  Sclerosis  Alliance)  Mike  Tranfaglia  (FRAXA)  Gerry  Fischbach  (Simons  Founda)on)  

Sage  TherapeuQcs  Gabi  Belfort  Carlos  Loya  Jim  Doherty  Rebecca  Hammond  

Seaside  TherapeuQcs  Paul  Wang  Randy  Carpenter  

GeneQcaLens/Well  Child  Lens  Louise  Tiranoff  Barbara  Floyd