Het Alzheimermysterie Wiesje M. van der Flier Workshop HA GNK 20 april 2017

Maakt u kennis met… Ben (53); Alzheimer patient. Op de fiets maakt hij nog altijd vele kilometers… …maar het herkennen van zijn fiets lukt niet meer.

Wist u dat… De ziekte van Alzheimer de allerduurste ziekte is (budget VWS: bijna 5 miljard!)? mantelzorgers heel hoog risico hebben op angst en depressie? de aantallen snel toenemen, terwijl de beroepsbevolking steeds kleiner wordt ( wie moet voor hen zorgen!)?

? Alzheimermysterie HOE kan het dat we van zo’n GROTE ziekte nog zo WEINIG weten Waarom krijgt iemand alzheimer? Of waarom niet (wie is beschermd)? Hoe ontstaat alzheimer? Hoe kun je alzheimer aantonen? Hoe kunnen we het genezen of voorkómen? oorsprong diagnostiek preventie

Dat komt doordat…. ! Geschiedenis alzheimer-onderzoek pas kort (~30 jaar) Investeringen in onderzoek bescheiden (vgl kanker, HvZ, HIV) Alzheimer=hersenziekte. Lastig te bestuderen  Al veel bereikt!

Het begint met een goede diagnose diagnostiek Het begint met een goede diagnose McKhann , Neurology 1984 1984: NINCDS-ADRDA criteria Memory impairment + Other cognitive impairment Decline from previous Impact on daily living Progressive Not due to other disorder Concept probable AD tijdens leven (Seniele) dementie vanaf nu ziekte Pathologische kenmerken geen onderdeel Diagnose per exclusionem

De uitdaging van de diagnose Diagnose gebaseerd op klinische criteria Ziekte veroorzaakt door hersenschade: (i) neuritische plaques, (ii) neurfibrillaire tangles  Microscopisch klein

Bij een gevorderde beeld.. .. kunnen de klinische criteria volstaan kloktekening: 23.10

Maar de uitdaging.. Wat als: Zorgen over geheugen, maar tests niet conclusief?

Ontwikkelingen in diagnostiek alzheimer biomarkers MRI krimp hele brein krimp stukken brein hersenvocht alzheimereiwitten PET alzheimereiwit ophopingen

NIA-AA diagnostische criteria McKhann, Alz&Dem 2011; Albert Alz&Dem 2011; Sperling Alz&Dem 2011 Alzheimer hersenschade is aanwezig vóór dementie Criteria voor Dementie MCI Preklinische AD Biomarker als bewijs voor AD pathofysiologie due to AD Biomarker bewijs te gebruiken om dementie with lage, medium, or hoge waarschijnlijkheid toe te schrijven aan AD “When available and deemed appropriate by the clinician” Richtlijnen voor hoe in te passen in klinische praktijk ontbreken..

When deemed necessary…?! Geen duidelijke aanbevelingen voor praktijk Geen enkele test perfect Optimale meetmethodes? Welke test wanneer en bij wie? Wat te doen bij conflicterende testresultaten? Abnormale biomarker  dementie Hoe leg je het allemaal uit aan de patient? En.. wat wil de patient eigenlijk? Het onderzoek ABIDE zoekt naar antwoorden!

…en dat is de grootste winst! Wat is bereikt: diagnose Vroeger: diagnose per exclusionem. Nu: inclusief Voorwaarde voor ontwikkelen therapie  Gereedschappen om het ziekteproces te bestuderen …en dat is de grootste winst! SCD-Initiative working group wil apply for PIA status.

Preklinische Alzheimer oorsprong Alzheimer ontstaat geleidelijk Dementie obv AD is slechts klinisch manifeste stadium langste deel ziekte speelt zich af vóór dementie (15-20 jaar) eerste hersenveranderingen milde geheugenproblemen dementie Figure 2 uit Bateman NEJM 2013. Comparison of Clinical, Cognitive, Structural, Metabolic, and Biochemical Changes as a Function of Estimated Years from Expected Symptom Onset. The normalized differences between mutation carriers and noncarriers are shown versus estimated years from expected symptom onset and plotted with a fitted curve. The order of differences suggests decreasing Aβ42 in the CSF (CSF Aβ42), followed by fibrillar Aβ deposition, then increased tau in the CSF (CSF tau), followed by hippocampal atrophy and hypometabolism, with cognitive and clinical changes (as measured by the Clinical Dementia Rating–Sum of Boxes [CDR-SOB]) occurring later. Mild dementia (CDR 1) occurred an average of 3.3 years before expected symptom onset. 95% confidence interval bands are shown in Figure S2 in the Supplementary Appendix. 15 à 20 jaar Preklinische Alzheimer

Amyloid bij gezonde ouderen: PET Villemagne, Ann Neurol 2011 PIB in 35 AD + 65 MCI + 106 gezonde controles PIB positief: AD 97%; MCI 69%; controles 31% PIB positieve controles: 25% MCI of AD na 3 jaar (1 PIB negatieve controle kreeg MCI) PIB scan = Amyloid PET-scan

Normale ouderen Vragen: 1) Hoe vaak komt preklinische Alzheimer voor? Vos, Lancet Neurol 2013 Vragen: 1) Hoe vaak komt preklinische Alzheimer voor? 2) Voorspelt preklinische Alzheimer achteruitgang? Methoden: 311 normale ouderen (>65 jaar) ruggenprik  alzheimerschade 5 jaar gevolgd Conclusie: 31% preklinische Alzheimer kans achteruitgang Figure uit Vos Lancet Neurol 2013: Figure: Progression to clinical dementia rating scale at least 0·5, symptomatic Alzheimer’s disease by preclinical Alzheimer’s disease stage Cumulative incidence probability of CDR at least 0·5, symptomatic AD for each preclinical AD stage (A) uncorrected for covariates and (B) corrected for age, sex, education, and APOE genotype. AD=Alzheimer’s disease. CDR=clinical dementia rating scale (range 0–3, with 0 as the best score). SNAP=suspected non-Alzheimer pathophysiology.

Vergeetachtigheid Patienten komen bij geheugenpolikliniek Van Harten, Neurology 2013 Patienten komen bij geheugenpolikliniek Voelen vergeetachtig, geheugentest normaal Vragen: 1) Hoe vaak komt preklinische Alzheimer voor? 2) Voorspelt preklinische Alzheimer achteruitgang? Methoden: 132 pp met subjectieve klachten ruggenprik  alzheimerschade 2 jaar gevolgd Conclusie: 16% preklinische Alzheimer kans achteruitgang geheugenprestatie Figuur: Van Harten, Neurology 2013

Iedereen met amyloid dementie? Wat weten we nog niet? geinspireerd op: Jack Lancet Neurol 2010 2. Verschillende routes? 1. Iedereen met amyloid dementie? Tijd tot dementie? 3.

Op zoek naar de oorsprong SCIENCE: Subjective Cognitive Impairment Cohort 300 vergeetachtige mensen Met gewone geheugentest Uitgebreide work-up baseline Amyloid-imaging Lange follow-up Samen proberen we meer te ontdekken over de vroegste hersenveranderingen And with these perhaps thought provoking ideas, I would like to briefly dwell on the next steps we have to take. Clearly, and I would like to re-iterate, SCD does not equal preclinical AD. We do not want to demonize. To answer some of these questions, we have set up the SCIENCe cohort in Amsterdam

Wat we zouden willen… preventie Dat is dus preventie! Preventie is ook: amyloid voorkómen

Op de goede weg? Barnes & Yaffe Lancet Neurol 2011; de Bruijn BMC Med 2015 Daling incidentie door betere behandeling hart- en vaatziekten? Rotterdam studie: alle modificeerbare risicofactoren uitbannen (!)  risico 23-33% 

Op de goede weg? Goed voor lijf = goed voor hersenen Barnes & Yaffe Lancet Neurol 2011; de Bruijn BMC Med 2015 Goed voor lijf = goed voor hersenen Westen  of  vs Niet-Westerse wereld  NB: Veranderingen leefstijl laatste 10 jaar? Voor de overige 70% moeten we onverminderd op zoek! Onderzoek loont!!

We staan pas aan het begin! Alzheimer: waar staan we nu We staan pas aan het begin! diagnose Diagnose = start inrichten zorg Voorwaarde ontwikkeling medicijn Gereedschap voor bestuderen ziekte oorsprong Hoe ontstaat alzheimer? Route 1e verandering  dementie? Oorsprong ziekte sleutel tot therapie preventie Voorkomen is beter dan genezen Zijn we al op de goede weg? (evt knoop in logo: om niet te vergeten… Netwerk nodig; vragen ontwarren) Alzheimer: de uitdaging voor de gezondheidszorg in de komende eeuw. We kijken aan tegen enorme aantallen. Een ziekte waarvoor geen geneesmiddel bestaat. De kosten – macro-economisch ‘en micro-economisch; financieel, maar ook sociaal en emotioneel, zijn enorm en zullen de komende decennia rap toenemen. Reden om bij de pakken neer te gaan zitten? Een onoplosbaar probleem? Wij denken van niet! Denk aan het voorbeeld van HIV-Aids Credo van het VUmc, en een uitgangspunt van het Alzheimercentrum: KENNIS MAAKT ONS BETER!

Meer lezen? Het Alzheimermysterie Het mooiste woord is herinnering Antwoorden op alle vragen die men ons keer op keer stelt Wat weten we al? Maar ook; wat we nog niet weten! En in welke richtingen we dan zoeken Het mooiste woord is herinnering Interviews met bekende Nederlanders over hun rol als mantelzorger Intieme portretten Prachtige foto’s

VUmc Alzheimercentrum

Phenotypical heterogeneity in Alzheimer disease: Effects of age-at-onset Wiesje M. van der Flier 4th Singapore International Neurocognitive Symposium March 27th 2015 25min incl vragen 26

Variation in phenotype Most important characteristic AD: memory loss But also: non-memory presentation extreme form: posterior cortical atrophy logopenic aphasia  Atypical variants memory PCA logopenic

Variation in phenotype Most important characteristic AD: memory loss But also: non-memory presentation But variation is not restricted to a few atypical cases memory executive function language visuospatial But I would like to make the case that this phenotypic variation is not restricted to a few, rare atypical cases. Rather, there is large, to often subtle variation in clinical manifestation. 28

Strongest risk factors Fratiglioni 200; Lobo 200; Farrer, JAMA 1997 Age: prevalence and incidence double every 5 years Apolipoproteine E (APOE): e4 is risk factor Meta-analysis ~6000 AD ~9000 controls --/-- e4/-- e4/e4 1 (reference) OR =3 OR =15

From risk factor to modificator AD can also develop at a younger age AD can also develop in the absence of APOE e4  Do the most important risk factors influence clinical manifestation as well??

Aim To find evidence for phenotypic heterogeneity in AD, in terms of: (genetic) make-up amyloid deposition neuronal injury neuro-degeneration clinical decline To find evidence for phenotypic heterogeneity in AD, in terms of: cognitive impairment brain structure brain function amyloid deposition In relation to age-at-onset In relation to age-at-onset and to APOE genotype, as examples of “what drives these possible differences”.

Amsterdam Dementia Cohort Patients visiting memory clinic VUmc Alzheimer center “Day screening”: One-stop shop Multidisciplinary Standardised Consent to use clinical data for research Routine work-up: Neurologist Spec nurse MRI EEG Neuropsych. Invest. Lab Lumbar puncture In relation to age-at-onset and to APOE genotype, as examples of “what drives these possible differences”.

Amsterdam Dementia Cohort Diagnosis by consensus in multidisciplinary meeting diagnosis postponed subjective complaints Other dementia Other In relation to age-at-onset and to APOE genotype, as examples of “what drives these possible differences”. ~600 new patients per year

Clinical variation exists, especially in early onset Clinical manifestation Koedam, JAD 2010 90 late onset + 270 early onset Review patient files Presenting complaint memory/non-memory Age-at-onset memory non-memory Presenting complaint: >65 yrs 65 yrs The clinical files of 270 patients with AD starting before the age of 65 years and 90 patients with late-onset AD ( 65 years) were reviewed to assess clinical characteristics. Patients were classified as memory presentation and non-memory presentation. The mean age of the early-onset group was 56 +/- 5 years and 74 +/- 6 years for the late-onset group. A third of the early-onset AD group presented with non-memory symptoms compared to only 6% in the late-onset group (p < 0.001). Within the group with non-memory presentations, apraxia/visuospatial dysfunction was the most prevalent presenting symptom (12%). Clinical variation exists, especially in early onset

Cognitive impairment: ~age Smits, JAD 2012 91 late onset + 81 early onset Neuropsych. test battery Domains; z-scores No difference MMSE (21±5) > 65 yrs  65 yrs memory language visuospatial exec.func attention worse better Early onset; different cognitive profile

Pattern of atrophy: MRI Typical MR characteristic atrophy medial temporal lobe

But in the young… ..often a different picture: Baseline 2 yr follow-up ..often a different picture: Remarkable posterior atrophy with spared hippocampus! Figure 1: MRI at baseline (left column) and at 2 years of follow-up (right column) in a woman aged 52 years with Alzheimer’s disease and APOE ε3/ε3 genotype A fl uid-attenuated inversion recovery sequence showed prominent posterior atrophy at baseline in the sagittal (A) and axial (C) views that progressed swiftly over 2 years (B, D). In the coronal view the hippocampus was not greatly aff ected at baseline (E), and the increase in atrophy was moderate at 2 years of follow-up, especially compared with the degree of posterior atrophy (F). 52-year old female with AD and APOE e3/e3 genotype

Voxel-based morphometry Moeller, NBA 2013 95 early onset + 120 late onset AD 97 younger + 32 older controls 3T MRI: T1-weighted VBM8 pipeline (SPM) Adjusted sex, TIV Family wise error correction LOAD < old controls EOAD < young controls Late onset: atrophy focused on (medial) temporal lobe Early onset: far more widespread atrophy in hippocampus, temporal lobes, precuneus, cingulate gyrus, and frontal cortex. Figure 1: MRI at baseline (left column) and at 2 years of follow-up (right column) in a woman aged 52 years with Alzheimer’s disease and APOE ε3/ε3 genotype A fl uid-attenuated inversion recovery sequence showed prominent posterior atrophy at baseline in the sagittal (A) and axial (C) views that progressed swiftly over 2 years (B, D). In the coronal view the hippocampus was not greatly aff ected at baseline (E), and the increase in atrophy was moderate at 2 years of follow-up, especially compared with the degree of posterior atrophy (F).

Voxel-based morphometry Moeller, NBA 2013 Direct comparison early vs late Adjusted sex, TIV, MMSE Family wise error correction EOAD < LOAD LOAD < EOAD Younger patients have more severe precuneus atrophy, despite their younger age

Function follows structure? Smits, Alz&Dem 2014 329 AD patients (67±8yr, 53%F, MMSE 20±5) MRI: visual rating scales; dichotomized medial temporal lobe atrophy posterior atrophy Neuropsychology: 5 domains MTA + PA MTA PA neither 26% 30% 17% 27%

Amyloid-beta: CSF Little relation with disease severity control Alzheimer amyloid-beta 1-42 young old Bouwman, NBA 2009 early onset + late onset: similarly low levels CSF amyloid Little relation with disease severity Little progression over time Amyloid: on/off phenomenon No meaningful variation? So how about amyloid? Literature shows... 41

Amyloid-beta: PIB-PET Rabinovici, Brain 2010 Little relation with disease severity Little progression over time Amyloid: on/off phenomenon No meaningful variation? 21 early onset + 18 late onset AD, 30 ctrls both early + late PIB uptake ~ cntrls No differences among AD groups Braak stages hold for all patients? 42

Amyloid-beta: PIB PET 45 younger AD + 46 older AD + 20 controls Ossenkoppele, Brain 2012 45 younger AD + 46 older AD + 20 controls Controls young AD old AD Young and old AD: same overall amyloid burden…

Younger patients relatively more parietal amyloid Amyloid-beta: PIB PET Ossenkoppele, Brain 2012 45 younger AD + 46 older AD + 20 controls Young and old AD: same overall amyloid burden… ..but difference in regional distribution (age*region) Younger patients relatively more parietal amyloid

Subtypes in Alzheimer’s disease? Typical AD memory old (~ 75 yrs) promoted by APOE e4 path: plaques&tangles CSF:ab42 + tau Amyloid-PET: positive atrophy: temporal slower progression Atypical AD non-memory young (~ 55 yrs) promoted by absence APOE e4 path: plaques&tangles CSF:ab42 + tau Amyloid-PET: positive atrophy:temporoparietal and/or frontoparietal faster progression

Where is the impact? Hierarchical ordered accumulation events in development AD genetic make-up amyloid deposition neuronal injury neuro-degeneration clinical decline

Where is the impact? genetic make-up amyloid deposition neuronal injury neuro-degeneration clinical decline Different subtypes; may be related to differences in genes and/or environmental factors  predispose for the same disease in terms of involved proteins, but infer specific regional vulnerability, which expresses as differences in changes in brain activity, brain structure and ultimately, clinical phenotype. (Genetic) variation seems to influence more downstream processes 47

Where is the impact? Also; subtle impact on upstream processes! genetic make-up amyloid deposition neuronal injury neuro-degeneration clinical decline Also; subtle impact on upstream processes! Different subtypes; may be related to differences in genes and/or environmental factors  predispose for the same disease in terms of involved proteins, but infer specific regional vulnerability, which expresses as differences in changes in brain activity, brain structure and ultimately, clinical phenotype. 48

Butterfly effect… genetic make-up amyloid deposition neuronal injury neuro-degeneration clinical decline Small effect upstream gradually increases to large effect downstream Butterfly; when a butterfly flaps its wings, this may cause a storm at the other side of the world. Or a drop in the water causes ever larger rings. Small effect upstream has far larger and more clear effect downstream. Alternative: some patients have a relatively higher burden of amyloid, needing less neuronal injury to result in clinical AD, whereas for others, this is the other way around. When a certain threshold of pathology is reached, clinical AD ensues. This may happen via a number of different routes. 49

Conclusion Meaningful phenotypical heterogeneity in AD not restricted to a few atypical cases evident in terms of cognitive profile pattern of atrophy brain activity amyloid deposition Can be linked to age-at-onset. Suggests that innate (unknown) factors result in partially different disease pathways Suggests that innate (unknown) factors result in partially different disease pathways Direct disease to different regions of the brain And/ or: make that specific parts of the cascade are more or less represented. For example relatively much amyloid and less neurodegeneration. Or vice versa (different disease aspects sum up to threshold) It is exactly these factors, that differ amongst patients and result in partially different disease pathways, that eventually may provide the key to finding therapeutic strategies that work for specific subsets of patients. 50

Meer lezen? Het Alzheimermysterie Het mooiste woord is herinnering Antwoorden op alle vragen die men ons keer op keer stelt Wat weten we al? Maar ook; wat we nog niet weten! En in welke richtingen we dan zoeken Het mooiste woord is herinnering Interviews met bekende Nederlanders over hun rol als mantelzorger Intieme portretten Prachtige foto’s

VUmc Alzheimercentrum