Biomarkers for Alzheimer s Disease Henrik Zetterberg, MD, PhD Professor of Neurochemistry The Sahlgrenska Academy, University of Gothenburg 1 Alzheimer s disease 2 Neuropathological criteria for Alzheimer s disease Senile plaques Neurofibrillary tangles General neuronal loss 3 Modified from: Blennow et al., Lancet26 1
Distribution of brain changes in Alzheimer s disease Mild cognitive impairment Mild dementia Moderate dementia Severe dementia 4 Molecular pathology Senile plaques are composed of a 42 amino acid long aggregation-prone protein called amyloid b 42 (Ab42) Neurofibrillary tangles are composed of hyperphosphorylated isoforms of the intraaxonal protein tau (P-tau) Neuroaxonal degeneration is reflected by release of intraaxonal tau proteins (all isoforms, total-tau; T-tau) into the brain interstitial fluid that communicates freely with CSF 5 Choosing the sample for biomarker analysis Brain CSF Blood Proximity to disease Accessibility Concentration of analyte Proteolytic degradation 6 2
Choroid Plexus Cerebrospinal fluid Choroid Plexus Circulation of CSF CSF in gray, Choroid Venous blood in orange Plexus Epidural vein 7 Total volume 15 ml, production rate 2 ml per hour Standard volume of sampling, 1-12 ml Is lumbar puncture dangerous? Consecutive studies on complications after LP No. of cases: 395 Post LP headache: 2.1 % Meningitis / hematoma: Blennow et al., 1993 No. of cases: 241 Post LP headache: 4.1% Meningitis / hematoma: Andreasen et al., 21 8 No. of cases: 342 (428 LP) Post LP headache:.9 % Meningitis / hematoma: Peskind et al., 25 No. of cases: 189 Post LP headache: 2.6 % Meningitis / hematoma: Zetterberg et al., Eur Neurol, 21 Core CSF biomarkers for Alzheimer s pathology Neurofibrillary tangles P-tau Axonal injury T-tau 9 Senile plaques Ab42 3
Biomarkers for Alzheimer s disease CSF biomarkers for Alzheimer s pathology - candidates Neurofibrillary tangles P-tau Axonal injury T-tau Oxidative stress F2-isoprostanes 1 Blood-brain barrier damage CSF/serum albumin ratio Inflammation Cell count IgG/IgM production a1-act, MCP-1, TNF-a, TGF-b Amyloid pathology Ab42 / Ab4 b-sapp / a-sapp BACE1 activity Different uses of biomarkers in the evaluation of patients with memory problems 1. To diagnose neuroinflammatory and neuroinfectious conditions CSF cell count Albumin ratio Intrathecal IgG and IgM production Specific inflammatory markers 2. Specific markers of AD neuropathology T-tau, P-tau and Ab42 as supporting diagnostic markers in the clinic T-tau, P-tau and Ab42 as additional inclusion criteria in studies of anti-ad drugs 3. Specific markers for drug effect monitoring Reduced levels of T-tau after treatment indicate less intense axonal degeneration 11 CSF biomarkers for AD: Ab1-42 b-secretase BACE1 a-secretase Plasma membrane g-secretase 12 4
CSF biomarkers for AD: Ab42 Study design: Consecutive AD cases from a community-based sample AD (n=53), healthy controls (n=21) 13 CSF-Ab42 25 Sensitivity 92% Specificity 95% 2 15 1 5 AD Controls Andreasen et al., Arch Neurol 56: 673, 1999 CSF Ab42 reflects plaque pathology Study design: 155 autopsy cases Plaque counts - Bielschowsky stain of neocortex and hippocampus Post-mortem CSF samples CSF-Ab42 (pg/ml) 14 8 6 4 2 Demented Non-demented 5 1 15 2 Plaque count Data adjusted for: Age at death Education Dementia severity ApoE4 Time from diagnosis until death PM interval Time until Ab42 analysis CAA severity Strozyk et al. Neurology 6: 652-656, 23 CSF Ab42 reflects plaque pathology (2) General idea verified in PIB-PET studies: Negative correlation between CSF-Ab 1-42 and PIB retention in the brain Posterior cingulum PIB (ROI/ref) Ab 1-42 15 MCI non-converters MCI converters Fagan AM et al., Ann Neurol 26 Forsberg A et al., Neurobiol Aging 28 5
Diagnostic value High total-tau and phospho-tau and low Ab1-42 in CSF = 85-95% sensitive and specific for AD in AD-control and longitudinal MCI studies >1 papers 16 Prediction of incipient AD in MCI cases using CSF biomarkers Study design: Follow-up study on MCI (>4 years) Lumbar puncture and analysis of Aβ42 and variants of tau in CSF MCI n=134 57 MCI AD 56 MCI MCI 21 MCI other dementias Comb. of Ab42/P-tau and T-tau: Sens. for MCI-AD 95% Spec. against other MCI 87% 17 Months Hansson et al., Lancet Neurol 26 Prediction of AD within 1 years in patients with MCI Cum survival Combination of Ab42/P-tau and T-tau: T-tau > 35 pg/ml Ab42 / P-tau ratio < 6.5 Sens. for MCI-AD - 88% Spec. non MCI-AD - 94% PPV - 94% NPV - 89% Time under risk for AD (months) 18 Buchhave et al., Arch Gen Psychiat212 6
CSF biomarkers for Alzheimer s disease diagnostic performance in a homogeneous mono-center population, 21 AD patients: n=32 Stable MCI: n=13 Other dementias: n=15 Controls: n=2 Sensitivity% 19 1% - Specificity% Johansson P et al., Journal of Alzheimer's Disease 24(3): 537-546, 211 Prediction of incipient AD in MCI cases using CSF biomarkers in a multi-center study MCI-AD vs. controls MCI-AD vs. stable MCI + MCI-other Ab 42/P-tau181 ratio Ab 42/P-tau181 ratio 2 CSF T-tau (ng/l) Sensitivity 83% Specificity 88% CSF T-tau (ng/l) Sensitivity 83% Specificity 72% Mattsson et al., JAMA 29 The Alzheimer s Association QC program for CSF biomarkers Goal 1 - Establish a standardized protocol for lumbar puncture and CSF processing Goal 2 - Monitor analytical variability Goal 3 - Establish a detailed protocol for laboratory procedures for AD CSF biomarkers Goal 4 - Improved quality and stability of assays 21 Please sign up if you perform Alzheimer biomarker analyses: Neurochem@neuro.gu.se 7
Can CSF markers predict development of AD in pre-symptomatic individuals? Study design: Population-based study, 55 healthy individuals LP at baseline, mean age 72 Follow-up 8 years later: 45 remained normal, 1 AD with dementia or drop in MMSE > -5 Baseline CSF Ab42 12 1 8 6 4 2 14 p =.6 r =.4 p<.5 Baseline CSF Ab42 12 1 8 6 4 22 Normal Dementia/ MMSE drop Normal Normal -12-9 -6-3 3 Change in MMSE Gustafson et al., JNNP 27 Can CSF markers predict development of AD in pre-symptomatic individuals? (2) Study design: Clinical study on 57 healthy elderly individuals LP at baseline Follow-up during 3 years No Yes No Yes No Yes Subjective memory impairment affecting quality of life (memqol) at follow-up 23 CSF Ab42 is lower in healthy elderly who will develop cognitive disurbances Stomrud et al., Dement Geriatr Cogn Disord 27; 24: 118-124 Can we measure Ab oligomers in CSF? Y Epitope blocked Y Abeta Abeta Y Abeta Y 24 Abeta = b-amyloid = biotin = avidin-hrp Y = 82E1, N-terminus-specific MAb = chemiluminescent substrate Xia W et al., Arch Neurol. 29 Feb; 66(2): 19-9 Fukumoto H et al., FASEB J. 21 Aug; 24(8): 2716-26 8
Can we measure Ab oligomers in CSF? (2) Oligomer in-house assay P<.1 Conc pg/ml Controls AD 25 Hölttä et al., in preparation Are CSF biomarkers dynamic? 26 Are CSF biomarkers dynamic? (2) Neurofilament Light Protein, ng/l Total Tau, ng/l Glial Fibrillary Acidic Protein, ng/l After a Bout (7-1 d) After Rest (3 mo) After a Bout (7-1 d) After Rest (3 mo) After a Bout (7-1 d) After Rest (3 mo) 27 Zetterberg et al.,arch Neurol 26 9
AD drug development 28 If you are interested in details of this, please read: Zetterberg, H. et al., (21) Alzheimers Res Ther 2(6): 32 29 Biomarkers in studies of inhibition of Ab aggregation 3 Change from Baseline at Week 12 Trial design: Randomized, double-blind, placebo-controlled, parallel study 12 weeks oral treatment with mg (placebo), 5mg and 25mg of PBT-2 CSF Ab 42 (pg/ml) CSF Ab 4 (pg/ml) CSF Ab 4 (pg/ml) CSF Ab42 (pg/ml) 1
Change from baseline (pg/ml) 31 CSF biomarkers in active Ab immunotherapy trials CSF substudy in the AN1792 trial (active β-amyloid immunization) Paired CSF samples (baseline and 1 year) from AD cases: antibody responders n=11, placebo n=1 Placebo CSF Ab42 Antibody responders Placebo CSF T-tau Antibody responders Reduction with treatment (in antibody reponders) on the downstream biomarker T-tau No clear effect on b-amyloid (Ab42) Change from baseline (pg/ml) Gilman S, et al.arch Neurol 25 CSF biomarkers in passive Ab immunotherapy trials Phase II bapineuzumab trials: CSF samples from 27 bapineuzumab and 19 placebo cases CSF samples taken at baseline and 1 year Change at Week 54 (pg/ml) 2 1-1 -2-3 CSF Ab1-42 Bapineuzumab Placebo Change a t Week 54 (pg/ml) 15 1 5 CSF Abx-42 CSF Ab1-4 6 p=.159 4 2-2 -4-6 Bapineuzumab Placebo Bapineuzumab Placebo Change at Week 54 (pg/ml) CSF T-Tau CSF P-Tau 32 Change at Week 54 (pg/ml) 1 5-5 -1-15 p=.35 Bapineuzumab Placebo Change at Week 54 (pg/ml) p=.27 5 p=.13-5 -1-15 Bapineuzumab Placebo Blennow, Zetterberg et al.arch Neurol 212 Phase III clinical trails on bapineuzumab Phase III bapineuzumab APOE e4 trial: Multicenter, randomized, double- blind, placebo-controlled trial Mild-moderate AD (MMSE 16-26) Bapineuzumab n=658 Placebo n=432 Main outcome: change in ADAS-Cog over 18 months 33 Sperling R, presented at EFNS 212 11
CSF biomarkers in the Phase III bapineuzumab trial Phase III bapineuzumab APOE e4 trial: Multicenter, randomized, double- blind, placebo-controlled trial Mild-moderate AD (MMSE 16-26) Bapineuzumab n=658 Placebo n=432 Secondary outcome: change in CSF phospho-tau over 18 months 34 Sperling R, presented at EFNS 212 The current situation in AD drug development: Promising biomarker changes but no clinical benefit Too little, too late? Studies on preclinical AD needed Are we missing something? 35 Summary CSF biomarkers can monitor the neuropathology of AD: T-tau is a marker of cortical axonal degeneration P-tau is a marker of neurofibrillary tangle pathology Ab1-42 is a marker of plaque pathology These markers have high diagnostic accuracy and have been implemented in the recently revised clinical criteria for Alzheimer s disease (http://www.alz.org/research/diagnostic_criteria/) Current research is aiming at establishing biomarkers for microglial activation and synapse loss 36 Biochemical markers together with neuroimaging and clinical evaluation allow for making a pre-dementia diagnosis of AD and can be used to detect biochemical treatment effects in clinical trials of anti-ad drug candidates 12
Thanks for listening! Clinical Neurochemistry Laboratory Sahlgrenska University Hospital Mölndal/Gothenburg, Sweden 37 Funding: The Swedish Research Council, Swedish Brain Power, JPND and the Alzheimer s Association 38 13