1. What Should we Learn from the POPular Study?
Neal Kleiman
Director Cardiac Catheterization Laboratory
Methodist DeBakey Heart Center
Professor of Medicine
Weill Medical College of Cornell University

2. The Three Principles
Platelet ‘activity’ studies suggest that in patients who are clopidogrel-dependent, measures of platelet activity predict subsequent events

3. If this is really the case, then therapeutic decisions might be made based on platelet ‘activity’ studies.
Increase clopidogrel dose
Use newer P2Y12 antagonists and accept increased risk of bleeding
Add cilostazol
Substitute surgery for PCI

4. Some studies are hard to do, some aren’t.
LTA requires tech, centrifuging, and nephelometic measurements.
VerifyNow requires no pipette and can be done at the bedside or in the CCL.
PFA-100 requires little time, but is a fairly bulky machine.
Plateletworks requires little time and is a whole blood CBC assay.
Measuring adhesion to collagen fibrils and subsequent aggregation is incredibly painful and results are difficult to standardize.

6. What Do Activated Platelets Do?
Activation
Prothrombin 1 4 Va
Thrombin
Vasoconstrictors +
Fibrinogen Xa 2 3
Mitogens +
Growth Factors
Aggregation
Adhesion
6-8-98

7. Evidence that pre mRNA is Spliced in Activated Circulating Platelets, i.e. IL-1 b is Synthesized De Novo
IL-1 b is a cytokine that upregulates many of the genes involved in inflammation
Denis, Cell:2005: 122: 379

8. Correlation Coefficients Between Assays
Lordkipanidze, M. et al. Eur Heart J :ehn419v1-9; doi: /eurheartj/ehn419

9. Gray lines represent 95% CIs for agreement
Bland-Altman Plots of Agreement Between Assays
Gray lines represent 95% CIs for agreement
k varied from 0.06 to 0.36
Lordkipanidze, M. et al. Eur Heart J : ; doi: /eurheartj/ehn419

10. Inter-individual Variability:
Clopidogrel 600 mg
100
y= * e-1.072* x
r2=0.101 80
n = 1001
600 mg Clopidogrel 60
Maximal Aggregation 5 µmol/L ADP (%) 40 20 2 4 6 8 10
Time from Loading Dose to Catheterization (hr)
Hochholzer W et al. Circulation. 2005;111: 10

11. 10% ↑ in aggregation  OR 1.32 (1.04-1.61) of 30 d MACE
Prospective Study of ADP Induced Aggregation in Patients Loaded with 600 mg of Clopidogrel
802 Patients; 19 events in d
Aggregation in response to 5 mM ADP
30 Day MACE (%)
↑ Age
↑ BMI
↑ % with Diabetes
↑ Active smokers
10% ↑ in aggregation  OR 1.32 ( ) of 30 d MACE
Hochholzer: JACC;2006: 48:1742

12. Effect of CYP2C19 Haplotypes on Ex vivo and Clinical Responses to Clopidogrel
Mega, NEJM: 2009;360:

13. U Mass Study of Cox-1 Dependent and Independent Markers of Platelet Activity
682 Patients presenting for diagnostic catheterization
34% also on clopidogrel
Followed for death/MI/revasc for months
Frelinger, A. L. et al. Circulation 2009;120:

14. POPular Study
Consecutive patients on clopidogrel undergoing elective stent implantation
N= (~2/3 DES)
LTA 5 mM ADP
n=1049
LTA 5 mM ADP
n=1051
VerifyNow P2Y12
n=1052
IMPACT-R
N=910
IMPACT-R ADP
N=910
PFA-100 COL/ADP
N=812
Primary Endpoint – 1 Y
Death, MI, Stent Thrombosis, or Stroke
Safety EP – 1 Y
Major and Minor Bleeding (TIMI grade)

15. POPular – Assays Tested
Light transmittance aggregometry
5 and 20 mM ADP in PRP
VerifyNow P2Y12 assay
Agglutination test
Plateletworks
Whole blood platelet count
IMPACT-R
Cone and platen viscometer w and w/o ADP
PFA-100
Shear stress (narrow aperture)

18. Cutoff Values Derived from ROC and 1 Year Outcomes
Cutoff Value (HPR vs NPR)
LTA 5 mM ADP
42.9%
LTA 20 mM ADP
64.5%
VerifyNow ADP
236 PRU
Plateletworks
80.5%
IMPACT-R
8.4 SC
IMPACT-R ADP
3.0 SC
PFA-100 COL/ADP
116 s
Innovance PFA
299 s

19. POPular—Prediction of the Composite Outcome
Logistic regression model incorporating clinical and procedural risk factors  AUC = 0.72
AUC P
LTA 5 mM ADP
0.74
0.004
LTA 20 mM ADP
0.73
<0.0001
VerifyNow P2Y12
Plateletworks
0.78
0.001
IMPACT-R
0.72
0.17
IMPACT-R -ADP
0.92
PFA-100
0.15
Innovance PFA P2Y
0.24

20. POPular -Bleeding
No assay predicted hemorrhage

21. What’s Required to WRAP the Package (1)?
Reproducibility
Prediction of a therapeutic resonse
Biologic plausibility
Are we measuring a function of platelet activity that affects the outcomes?
What is the temporal stability of the estimates?
Is the estimate independent of clinical status?
Is a single measurement of platelet ‘activity’ likely to predict an event that occurs one year later?

22. Reported Drug-Drug Interactions with Antiplatelet Regimens
Mechanism
Rantidine
Aspirin ↓
↓ Absorption
Atorvastatin
Clopidogrel
↓ CYP3A4
Omeprazole
↓ CYP2C19
Caffeine ↑
↑ cAMP level
Ibuprofen
Blockade of Ser529
Ca++ Block.
Rifampin
↑ CYP3A4

23. Temporal Variation in Aspirin Resistance Among Survivors of Suspected Acute MI
ASA Resistance
Baseline N=46
1 Year F/U N=17
AMI (n=37)
43%
(95% CI: 27% - 59%)
11%
(95% CI 1% - 21%)
No AMI (n=150)
20%
(95% CI 14%-26%) 9%
(95% CI 5% - 13%)
Poulsen: Thrombosis Research: 2007: 120: 477

24. What Should we Learn from POPULAR?
Confirmation of an association between high platelet reactivity and common ischemic events.
Attempt to establish appropriate cutoff points for each test.
Suggestion that platelet reactivity measurements may not be so independent of traditional clinical and procedural markers.

25. What Have We Yet to Learn from POPular? Remaining Needs
Validation of the determined cutpoints in another dataset.
To determine whether prediction of HPR with one assay also predicts HPR with other assays. Are they identifying the same patients?
Need to demonstrate a benefit for therapy guided by the assays.