Bioanalytics Core Update - Parkinson`s Progression Markers Initiative

Bioanalytics Core Update - Parkinson`s Progression Markers Initiative

Bioanalytics Core
Leslie M Shaw
PPMI Annual Investigators Meeting, NYC
May 8, 2013
11/16/2011
1
Bioanalytics Core-2013
• Pilot study manuscript accepted for publication in
•
•
JAMA Neurology
Preparations for analyses of CSF Aβ1-42, t-tau and ptau181 in PPMI baseline samples, and baseline, 6 and
12 month longitudinal samples using validated xMAP
AlzBio3 immunoassay (all longitudinal samples on
same 96 well plate
Aliquot comparisons; compare current lot of
immunoassay reagents with that used in 2012
11/16/2011
2
Association of cerebrospinal fluid Ab1-42, t-tau, p-tau181 and alpha-synuclein levels
with clinical features of early drug naïve Parkinson’s disease patients; a crosssectional study
By
Ju-Hee Kang, David J Irwin, Alice S Chen-Plotkin, Andrew Siderowf, Chelsea
Caspell, Christopher S Coffey, Teresa Waligórska, Peggy Taylor, Sarah Pan, Mark
Frasier, Kenneth Marek, Karl Kieburtz, Danna Jennings, Tanya Simuni, Caroline
M Tanner, Andrew Singleton, Arthur W Toga, Sohini Chowdhury, Brit
Mollenhauer, John Q Trojanowski, Leslie M Shaw and the Parkinson’s Progression
Marker Initiative
3
Transfer of PPMI CSFs
(Initial 102 CSF samples for statistical analysis)
15 Clinical Centers
Frozen CSFs
5 F/U CSFs
44 Controls
39 BL
HC CSFs
120 CSF samples
63 BL
PD CSFs
76 Patients
102 CSFs
for statistical
analysis
4 BL SWEDD CSFs
9 F/U CSFs
4
Association of cerebrospinal fluid Aβ1-42, t-tau, p-tau181 and α-synuclein
levels with clinical features of early drug naïve Parkinson’s disease patients
J-H Kang, DJ Irwin, AS Chen-Plotkin, A Siderowf, C Caspell, CS Coffey, T Waligórska,
P Taylor, S Pan, M Frasier, K Marek, K Kieburtz, D Jennings, T Simuni, CM Tanner,
A Singleton, AW Toga, S Chowdhury, B Mollenhauer, JQ Trojanowski, LM Shaw and the
Parkinson’s Progression Marker Initiative*
Objective: Evaluate baseline characteristics and relationship to clinical
features of CSF Aβ1-42, t-tau, p-tau181 and α-SYN in PD patients and matched healthy
controls (HC) enrolled in PPMI.
Methods: CSF biomarkers were measured by xMAP-Luminex platform and ELISA in
HC (N=39) and PD (N=63).
Accessioning and statistical analyses of PPMI
CSF biomarker dataset
•
•
•
•
•
•
Simultaneous download of PPMI data for the first 106 study subjects
6/30/2012 by IU and UPenn
Agreed upon a statistical analysis plan
Statistical analyses done at UPenn, SAS script sent to IU for replication
All results in concordance following this process and exchanges of
detailed analyses
Data assembly, incorporation into draft manuscript
Manuscript circulated amongst the primary authors and the study site
investigator/authors for suggestions, edits, further discussions of the
data
6
Statistical Analyses
• Software: SAS ver. 9.3
• Group comparison
— Mann-Whitney U test (2 groups)
—
Kruskal-Wallis test with Dunn’s multiple comparison (3 groups)
• Association of CSF biomarker levels and clinical variable
— Multiple linear or logistic regression model (stepwise selection)
with adjustment for confounders (Age, Gender and Education)
• Chi-square test for difference in percentage of subjects
• Pearson correlation
Demographics of the PPMI subjects
HC (N = 39)
PD (N = 63)
P value
Age, years (95% C.I.)
58 ± 13 (54 – 63)
62 ± 10 (59 – 64)
0.2390*
Sex, F/M (% of Male)
18/21 (53.8)
24/39 (61.9)
0.4216#
16.8 ± 2.4
16.4 ± 2.5
(16.0 – 17.6)
(15.8 – 17.0)
Age at diagnosis, years
(95% C.I.)
-
61.1 ± 10.0 (58.6 –
63.7)
-
Disease duration, median years
(range)
-
0.4 (0.0 – 2.6)
-
Number of subjects with CSF
Hgb > 200 ng/mL
6
18
0.1271#
Education, years (95% C.I.)
*Mann-Whitney U test; #Chi-square test for HC vs. PD.
0.1517*
Clinical characteristics of the PPMI subjects#
HC (N = 39)
PD (N = 63)
p value*
0.03 ± 0.16
1.65 ± 0.51
< 0.0001
1.6 ± 2.7
22.6 ± 7.6
< 0.0001
Mean tremor score
0.05 ± 0.13
0.53 ± 0.32
< 0.0001
Mean PIGD score
0.01 ± 0.04
0.24 ± 0.26
< 0.0001
UPSIT score
35.1 ± 3.4
21.9 ± 8.1
< 0.0001
PR¶/PL/CR/CL
PR/PL/CR/CL
1.38/1.39/2.06/2.05
0.62/0.64/1.35/1.34
MoCA (95% C.I.)
28.4 ±1.0 (28.0 –28.7)
27.2 ± 2.0 (26.7 – 27.7)
0.0054
Semantic fluency
53.8 ± 12.1
49.5 ± 10.6
0.0565
WMSIII-LNS$ test score
12.1 ± 2.8
11.0 ± 2.0
0.0510
SDMT
48.6 ± 11.2
41.9 ± 8.9
0.0051
HVLT-R total recall
9.0 ± 1.6
8.2 ± 1.5
0.0077
HVLT-R delayed recall
9.9 ± 2.3
8.3 ± 2.3
0.0004
H & Y stage
UPDRS III motor score
Striatal binding ratios
(Mean values)
#Data were updated based on PPMI database (06.30.2012)
*Mann-Whitney U test
¶PR: Right putamen, PL: Left putamen, CR: Right caudate, CL: Left caudate, N=39 for HC, N=62 for PD
<0.0001
Comparison of CSF Biomarker levels between HC and PD #
Aβ1-42 (pg/mL)
t-tau (pg/mL)
p-tau181 (pg/mL)
t-tau/Aβ1-42 ratio
p-tau181/Aβ1-42 ratio
p-tau181/t-tau ratio
α-syn (pg/mL)
*Mean
HC (N = 39)
PD (N = 63)
242.8 ± 49.95
228.7 ± 45.63
(226.7 – 259.0)*
(217.2 – 240.2)
53.9 ± 19.33
46.1 ± 24.71
(47.6 – 60.1)
(39.8 – 52.3)
24.9 ± 8.45
21.0 ± 7.83
(22.2 – 27.6)
(19.0 – 23.0)
0.240 ± 0.141
0.215 ± 0.157
(0.195 – 0.286)
(0.176 – 0.255)
0.113 ± 0.075
0.099 ± 0.063
(0.089 – 0.138)
(0.084 – 0.115)
0.491 ± 0.160
0.543 ± 0.263
(0.439 – 0.543)
(0.477 – 0.609)
1264 ± 425.7
1082 ± 611.1
(1126 – 1403)
(928 – 1235)
± S.D. (95% confidence interval); #Mann-Whitney U test.
#Data were updated based on PPMI database (06.30.2012)
P value#
0.0466
0.0276
0.0093
0.0451
0.1482
0.6820
0.0120
MDS-UPDRS Subsection used to classify
Tremor or PIDG-dominant phenotype
 Mean tremor score (11 items)
: UPDRS II – 1) Tremor
: UPDRS III – 2, 3) Postural tremor (both upper extremities), 4, 5) Kinetic tremor (both upper extremities), 6-10)
Resting tremor (4 extremities and lip/jaw), 11) Rest constancy
 Mean Postural Instability & Gait Disturbance (PIGD) score (5 items)
: UPDRS II – 1) Walking and balance, 2) Freezing
: UPDRS III – 3) Gait, 4) Freezing of gait, 5) Postural stability
Tremor dominant (TD), or PIGD dominant phenotype
- Ratio of tremor/PIGD score ≥ 1.15: Tremor dominant
- Ratio of tremor/PIGD score ≤ 0.90: PIGD dominant
- 0.90 < Ratio of tremor/PIGD score < 1.15: Intermediate type
- If PIGD score is 0, but tremor score > 0: Tremor dominant
CSF biomarkers according to clinical
phenotype in PD patients
HC
IND-PD
(N =39)
(N=6)
0.0323
242.8 ± 50.0
215.5 ± 25.0
50.3 ± 24.01
0.0527
53.9 ± 19.33
31.2 ± 9.97
18.0 ± 6.74#
22.5 ± 8.17
0.0387
24.9 ± 8.45
17.7 ± 4.97
α-syn (pg/mL)a
892.8 ± 542.4#
1185 ± 649.6
0.0587
1264 ± 425.7
782.6 ±150.1
α-syn (pg/mL)b
766.3 ± 446.3#
1122 ± 451.8
0.0286
1267 ± 443.5
775.9 ± 184.8
t-tau/Aβ1-42 ratio
0.211 ± 0.213
0.225 ± 0.145
0.1089
0.240 ± 0.141
0.151 ± 0.072
p-tau/Aβ1-42 ratio
0.093 ± 0.059
0.104 ± 0.068
0.2247
0.113 ± 0.075
0.083 ± 0.026
p-tau/t-tau ratio
0.617 ± 0.398
0.513 ± 0.217
0.7597
0.491 ± 0.160
0.588 ± 0.164
PIGD-PD
TD-PD
(N=14)
(N=43)
Aβ1-42 (pg/mL)
211.4 ± 45.0#
236.2 ± 46.8
t-tau (pg/mL)
39.3 ± 28.27#
p-tau181 (pg/mL)
Biomarkers
p
value*
*PIGD vs. TD; Mann-Whitney U test
# P<0.05 versus HC by Kruskal-Wallis test with Dunn’s multiple comparison
aα-syn
was measured in total subjects, or bsubjects with CSF hemoglobin level of < 200 ng/mL
12
Summary of multivariate regression analyses
•
•
•
Multivariate regression analysis: lower Aβ1-42 (p=0.0383) and p-tau181 (p=0.0015) are significantly
associated with PD diagnosis, but other biomarkers or their ratios are not.
For clinical variables in PD, α-syn (p=0.0081) was significantly associated with the MDS-UPDRS III
motor score and t-tau was marginally associated (p=0.0424).
We found that lower levels of CSF Aβ1-42 and p-tau181 were significantly associated with a higher PIGD
risk.
Correlation between AD biomarkers and α-synuclein:
Total
t-tau
p-tau181
Aβ1-42
PD
HC
Association of cerebrospinal fluid Aβ1-42, t-tau, p-tau181 and α-synuclein
levels with clinical features of early drug naïve Parkinson’s disease patients
J-H Kang, DJ Irwin, AS Chen-Plotkin, A Siderowf, C Caspell, CS Coffey, T Waligórska,
P Taylor, S Pan, M Frasier, K Marek, K Kieburtz, D Jennings, T Simuni, CM Tanner,
A Singleton, AW Toga, S Chowdhury, B Mollenhauer, JQ Trojanowski, LM Shaw and the
Parkinson’s Progression Marker Initiative*
Results: Significantly lower concentrations of all measured CSF biomarkers and t-tau/
Aβ1-42 ratio were seen in PD compared to HC, lower α-SYN was associated
with a higher risk of PD and decreased CSF p-tau181 associated with increased
UPDRS motor score. Notably, lower CSF Aβ1-42 was associated with the postural
instability-gait disturbance-dominant phenotype which associates with a more
rapid cognitive decline and poor prognosis compared to tremor-dominant patients. There is
a significant correlation between α-SYN and t-tau & p-tau in PD and HC subjects.
Interpretation: We demonstrate that CSF Aβ1-42, t-tau, p-tau181 and α-SYN have
value for diagnosis and assessment of disease progression in early-stage PD.
Further investigations will test the predictive performance of CSF biomarkers for disease
progression.
Bioanalytics Core 2013
• Analyses of the entire PPMI BASELINE CSFs, n=645
•
•
•
and CSFs for those 223 subjects whose 6 month & 12
month CSFs are collected
AlzBio3 immunoassay for Aβ1-42, t-tau and p-tau181 at
UPenn and α-SYN by ELISA at Covance
Data analyses, qc, data upload expected by mid to
late summer, 2013.
Statistical analyses will test hypotheses based on the
pilot study
Acknowledgements: John Trojanowski and the Penn PPMI Bioanalytics Core
members Ju Hee Kang, Teresa Waligorska and Sarah Pan, our PPMI study participants,
and the amazing PPMI Team!
PPMI Funding Partners
PPMI is sponsored and partially funded by The Michael J. Fox Foundation for
Parkinson’s Research. Other funding partners include a consortium of industry players,
non-profit organizations and private individuals.
17
Scatter plots of CSF biomarker levels in PD
vs. HC
α-syn
Aβ1-42
t-tau
p-tau181
Ratios