Pharmacologic Treatment of Dementia: A Critical Appraisal Including cholinesterase inhibitors, NMDA partial antagonists, antidepressants, and antipsychotics

Prevalence and Treatment Rates MildModerateSevere Number of Patients (thousands) Prevalence 1 Diagnosed 2 Treated with AChEI 3 Sources: 1. Hebert LE, Scherr PA, Bienias J, et al. Arch Neurol. 2003;60: Datamonitor AD Treatment Algorithms Market Measures

The Controversy …”doneppezil was not effective, … oversimplification could have unwarranted promotional value… Finucane TE. N Eng J Med 2004;35:1912 …”nearly all organizations [AAN, AAGP, EFNS] that have provided an opinion… have recommended their use”… Cummings JL. N Eng J Med 2004;35:1912

The Controversy …”strategy to sell Alzheimer’s disease (AD) drugs is based on… guilt if family members have not made the decision to “fight” the disease”… Worst Pills Best Pills News. Sidney M. Wolfe MD Editor. July 2004;10(7):49-50 “DO NOT USE — Memantine (Namenda)…common adverse effects ” Worst Pills Best Pills News. Sidney M. Wolfe MD Editor. July 2004;10(7):49-50

1 March 2005 National Institute for Clinical Excellence, an independent advisory agency of the UK’s National Health Service, preliminary recommendations, comments due 22 Mar 2005 Donepezil, rivastigmine, galantamine, memantine not recommended Each judged beneficial for cognition or global improvement None judged worthy of the expenditure Existing studies biased in selection, measurement, and attrition accessed 4 March 2005

Long-term donepezil treatment of 565 patients with Alzheimer’s disease (AD2000): randomized double-blind trial Cortney et al. Lancet 2004; 34: weeks donepezil=282, placebo=283 then re-randomize 48 weeks donepezil 5/10mg=125, placebo=283 then 6 week wash out 48 weeks donepezil=105, placebo=89 then 4 week washout 48 weeks donepezil=31, placebo=20 then 4 week washout 48 weeks donepezil=12, placebo = 8

AD2000 Findings-1 Institutionalization 42% vs 44% at 3 yrs ns Progression of disability 58% vs 59% at 3 yrs ns MMSE 0.8 point better 95% CI 0.5 — 1.2: p< Bristol ADL 1.0 point better 95% CI 0.5 — 1.6 p<0.0001

AD2000 Findings-2 Behavioral and psychological symptoms NPI 0.3 points better 95% CI -0.9 —1.5 ns Caregiver psychopathology GHQ 0.3 points better 95% CI -0.3 — 0.9 ns Formal care costs £ 2842 — £ 2344 ns Serious adverse events & deaths donepezil = 63 placebo = 50 ns

Schneider’s Commentary Lancet 2004;363: Strengths –Publicly, not industry, funded –Typical patients, typical practitioners –Demonstrates benefits beyond 12 weeks –Interrupted treatment not harmful Weaknesses –Under-powered to prove that nursing home admission may be delayed –Referral bias, MD bias, no switching

Reviews of Cholinesterase Inhibitors Trinh N-H, et al. Efficacy of cholinesterase inhibitors in the treatment of neuropsychiatric symptoms and functional impairment in Alzheimer’s disease: a meta- analysis. JAMA 2003;289: Santaguida PS, et al. Pharmacologic treatment of dementia. Summary, Evidence Report/Technology Assessment No. 97. AHRQ Publication No. 04E Rockville, MD: Agency for Healthcare Research and Quality. April 2004 Langa KM et al. Mixed dementia: Emerging concepts and therapeutic implications JAMA 2004;292: Birks JS, Harvey R. Donepezil for dementia due to Alzheimer’s disease. (Cochrane Review). In: The Cocrhane Library, Issue 2, Chichester: John Wiley & Sons Ltd, CD001190

Population-Based Study of Nursing Home Admissions and use of CEI Insurance database of 5562 patients with Alzheimer’s disease, 517 treated with either donepezil or rivastigmine New users only Untreated patients were almost 3 times more likely to be admitted over 2 years (p<0.001 adjusted hazard ratio = % CI ) Beusterien. Presented at the 2004 Annual ASCP meeting

Alteration of a clinically meaningful outcome in the natural history of Alzheimer’s disease by cholinesterase inhibition Lopez et al. J Am Geriatr Soc 2005;53:83-87 Rate of Change in MMSE at 1 Year CEI Placebo 3 pts 3 pts 81 (60%) 54 (40%) 53 (39%) 82 (61%) 12 mo MMSE Nursing home admission at 3 years 3 (4%) 4 (7%) 24 (29%) 18 (34%)

Patients receiving a cholinesterase inhibitor experienced…  Likelihood of being a slow progressor  RR = 2.32, 95% CI =  Progression after stepwise regression analysis  RR = 2.45, 95% CI =  Nursing home admission  RR = 0.15, 95% CI = Lopez et al. J Am Geriatr Soc 2005;53:83-87

Cumulative Nursing Home Admissions Over 3 Years Lopez et al. J Am Geriatr Soc 2005;53:83-87 placebo = 42/84 cholinesterase inhibitor = 6/56 p<0.001

Mean change from baseline in ADCS/IADL n=659 Galasko et al. JAGS 2004;52: IADL PlaceboGal 16Gal 24 0  p< P<

Patients Declining May Benefit from a Change Auricombe et al, Curr Med Res Opin, 2003 MMSE Mean score Start donepezil n=205 Stop denepezil Rivastigmine 26 weeks later p<.001 from start p<.001 from stop

Lack of Tolerability or Benefit from One Agent Does not Preclude Another Auricombe et al, Curr Med Res Opin, 2003 d/c RIV due to AEs RIV responders d/c RIV due to AEs RIV responders DON not tolerable DON failed 54% 74% 11%  15% 

Mild Cognitive Impairment Defined by –Memory complaints –Objective evidence of memory impairment –Not disabling –Other cognitive domains intact Non-amnestic MCI 15% per year conversion to dementia

A Randomized Double Blind Placebo Controlled Trial of Donepezil or Vitamin E to Delay the Conversion of MCI to Dementia Days to conversion to dementia (n=769) –Donepezil 10 mg 661 –Vitamin E 2000 IU 540 –Placebo 484 No significant difference between donepezil, vitamin E, and placebo at 36 months

How Do We Agree on Improvement? FDA –ADAS/COG CIGI or CIBIC+ Langa et al JAMA 2004;292:2901 –What individuals and families hope to achieve Royall JAGS 2005;53:163 –Expert bodies should promulgate standards; industry and the regulators strive to meet them during drug development –Clinician consensus Thies WK Alzheimer’s Association –“families and individuals …can decide with their physicians” (accessed 4 March 2005)

“Rose’s Theorem” – “...a large number of people at small risk may give rise to more cases of disease than a small number who are at high risk” (Rose 1989)

Donepezil ADAS-cog Scores After Washout Burns A et al. Dement Geriatr Cogn Discord. 1999;10: Week Placebo washout 0 Endpoint **** Placebo Donepezil 5 mg/day Donepezil 10 mg/day Mean change (± SE) from baseline in ADAS-cog scores Improvement ** **** *** **** *** * * p < ** p < *** p < **** p <

Efficacy of Rivastigmine on Cognition Through 52 Weeks *Alzheimer’s Disease Assessment Scale–Cognitive Subscale. † P<0.05 vs projected placebo. Farlow M, et al. European Neurology Weeks All Patients Taking Rivastigmine 2-12 mg/d † Improve Decline Baseline mg/d 1-4 mg/d Placebo Projected placebo (n=196) Mean Change from Baseline † † † † † † † †

Open-extensionDouble-blind Long-term Cognitive Benefits of Galantamine Treatment * p < 0.05 vs placebo/galantamine (not statistically different from baseline). Improvement Deterioration Mean (± SE) change from baseline in ADAS-cog score –4 –3 –2 – Months Galantamine 24 mg/galantamine Historical placebo group Placebo/galantamine 24 mg * Reprinted with permission from Raskind MA et al. Neurology. 2000;54: Torfs K et al. World Alzheimer Congress, 2000.

If Behavioral Interventions Fail Sink et al JAMA 2005;293:596 CEI +/- memantine  Anxiety, depression? SSRI   Atypical antipsychotic  SSRI  Carbamazepine

Galantamine 24 mg/d Placebo Galantamine 16 mg/d * p < 0.05 vs placebo galantamine 16 & 24 mg † p < 0.05 vs baseline. Improvement Deterioration –3 –2 – * Mean (± SE) change from baseline in NPI Tariot PN et al. Neurology. 2000;54: Months Dose increments † Effect of Galantamine on Behavioral Symptoms: NPI

Rivastigmine Lewy Body Trial: Behavioral Changes :NPI Baseline Week 12 Week 20 Rivastigmine Placebo Improvement *P<0.01 vs placebo (ANOVA/ANCOVA) McKeith IG, et al. American Academy of Neurology 52 nd Annual Meeting. April 29-May 6, San Diego, California. N/P-5 *

Results: Behavior—NPI At End Point There was no statistically significant difference between the 2 groups for total NPI scores There was a statistically significant difference between the treatment groups in favor of memantine in the following domains –Delusions P =.0386* –Agitation/aggression P =.0083* Memantine in Moderate to Severe AD Study *LOCF analysis. Sources: Reisberg B, et al. N Engl J Med. 2003;348: Data on file, Forest Laboratories, Inc.

Antidepressant Studies in Dementia Author Yr N Medications Outcome Reifler Imipramine/Plac No difference Nyth Citalopram/Plac No difference Nyth Citalopram/Plac CIT > 1 sx Petracca Chlorimipr/Plac CMI > Plac Roth Moclobem/Plac MOC > Plac Lyketsos & Olin. Biol Psychiatry 2002;52:243

Antidepressant Studies in Dementia Author Yr N Medications Outcome Taragano Fluox/Amitrip 78 vs. 42% Katona Imipram/Paroxet No difference Magai Sertraline/Plac No difference Lyketsos Sertraline/Plac Sert > Plac Lyketsos Sertraline/Plac Sert > Plac Petraca Fluoxitine/Plac No difference Lyketsos & Olin. Biol Psychiatry 2002;52:243

Risperidone in Dementia (cont) Psychosis Subscale —BEHAVE- AD Mean shift at endpoint Mean Improvement From Baseline * *P .005 vs placebo. † P .01 vs placebo. Source: Katz et al. J Clin Psychiatry. 1999;60: Dose of risperidone † 2.0 mg n = 162 Placebo n = mg n = mg n = 148

* ** Key Efficacy Assessments: Change from Baseline to Week 10 LOCF Change from Baseline Various Efficacy Outcome Measures *p=0.023 QTP vs. Pbo; **p=0.005 QTP vs. HAL; † NPI-NH2 = sum of hallucinations and delusions Tariot P et al, 2004 (Manuscript under review) Quetiapine Haloperidol Placebo BPRS BPRS BPRS NPI-NH2 † NPI-CGI-S TotalAgitationAnergia Agitation

Aripiprazole for Psychosis of AD (Trial 3): Mean Change in CMAI *P<0.05 vs placebo. CMAI = Cohen-Mansfield Agitation Inventory, scale Breder et al. APA, * Weeks Placebo (n=115, baseline = 55.6) Aripiprazole 10 mg/d (n=118, baseline = 54.2) X Aripiprazole 2 mg/d (n=114, baseline = 57.4) X Aripiprazole 5 mg/d (n=115, baseline = 58.7) * * * * * CONTAINS OFF-LABEL INFORMATION

Summary Pharmacotherapy for the impaired cognition as well as the psychological and behavioral symptoms of dementia are modestly but reliably effective for the target symptom. A comprehensive approach combining behavioral and pharmacologic approaches is widely accepted but data to demonstrate the social significance is lacking. Both clinical practice and public expectations have outstripped the evidence base. Uncertainty about meaningful outcomes can be reduced with studies of new interventions.

ADL PlaceboGal 16Gal 24 0  P< P<0.05 Mean change from baseline in ADCS/ADL n=659 Galasko et al. JAGS 2004;52: